diff --git a/generation/maisi/README.md b/generation/maisi/README.md index 727b1a2667..5c596b75d2 100644 --- a/generation/maisi/README.md +++ b/generation/maisi/README.md @@ -2,14 +2,25 @@ This example demonstrates the applications of training and validating NVIDIA MAISI, a 3D Latent Diffusion Model (LDM) capable of generating large CT images accompanied by corresponding segmentation masks. It supports variable volume size and voxel spacing and allows for the precise control of organ/tumor size. ## MAISI Model Highlight +**Initial Version (August 2024):** First release `maisi3d-ddpm`. + - A Foundation Variational Auto-Encoder (VAE) model for latent feature compression that works for both CT and MRI with flexible volume size and voxel size. Tensor parallel is included to reduce GPU memory usage. - A Foundation Diffusion model that can generate large CT volumes up to 512 × 512 × 768 size, with flexible volume size and voxel size - A ControlNet to generate image/mask pairs that can improve downstream tasks, with controllable organ/tumor size More details can be found in our WACV 2025 paper: -[Guo, P., Zhao, C., Yang, D., Xu, Z., Nath, V., Tang, Y., ... & Xu, D. (2024). MAISI: Medical AI for Synthetic Imaging. arXiv preprint arXiv:2409.11169](https://arxiv.org/pdf/2409.11169) -Welcome to try our GUI demo at [https://build.nvidia.com/nvidia/maisi](https://build.nvidia.com/nvidia/maisi). +[Guo, P., Zhao, C., Yang, D., Xu, Z., Nath, V., Tang, Y., ... & Xu, D. (2024). MAISI: Medical AI for Synthetic Imaging. WACV 2025](https://arxiv.org/pdf/2409.11169) + +πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰**Release Note (March 2025):** πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰πŸŽ‰ + +We are excited to announce the new MAISI Version `maisi3d-rflow`. Compared with the previous version `maisi3d-ddpm`, **it accelerated latent diffusion model inference by 33x**. The MAISI VAE is not changed. The differences are: +- The maisi version `maisi3d-ddpm` uses basic noise scheduler DDPM. `maisi3d-rflow` uses Rectified Flow scheduler. The diffusion model inference can be 33 times faster. +- The maisi version `maisi3d-ddpm` requires training images to be labeled with body regions (`"top_region_index"` and `"bottom_region_index"`), while `maisi3d-rflow` does not have such requirement. In other words, it is easier to prepare training data for `maisi3d-rflow`. +- For the released model weights, `maisi3d-rflow` can generate images with better quality for head region and small output volumes than `maisi3d-ddpm`; they have comparable quality for other cases. +- `maisi3d-rflow` added a diffusion model input `modality`, which gives it flexibility to extend to other modalities. Currently it is set as always equal to 1 since this version only supports CT generation. We predefined some modalities in [./configs/modality_mapping.json](./configs/modality_mapping.json). + +**GUI demo:** Welcome to try our GUI demo at [https://build.nvidia.com/nvidia/maisi](https://build.nvidia.com/nvidia/maisi). The GUI is only a demo for toy examples. This Github repo is the full version. @@ -29,7 +40,8 @@ We retrained several state-of-the-art diffusion model-based methods using our da | [DDPM](https://proceedings.neurips.cc/paper_files/paper/2020/file/4c5bcfec8584af0d967f1ab10179ca4b-Paper.pdf) | 18.524 | 23.696 | 25.604 | 22.608 | | [LDM](https://openaccess.thecvf.com/content/CVPR2022/papers/Rombach_High-Resolution_Image_Synthesis_With_Latent_Diffusion_Models_CVPR_2022_paper.pdf) | 16.853 | 10.191 | 10.093 | 12.379 | | [HA-GAN](https://ieeexplore.ieee.org/document/9770375) | 17.432 | 10.266 | 13.572 | 13.757 | -| MAISI | 3.301 | 5.838 | 9.109 | 6.083 | +| MAISI (`maisi3d-ddpm`) | 3.301 | 5.838 | 9.109 | 6.083 | +| MAISI (`maisi3d-rflow`) | 2.685 | 4.723 | 7.963 | 5.124 | **Table 1.** Comparison of FrΓ©chet Inception Distance scores between our foundation model and retrained baseline methods
using the unseen public dataset [autoPET 2023](https://www.nature.com/articles/s41597-022-01718-3) as the reference. @@ -39,7 +51,7 @@ We retrained several state-of-the-art diffusion model-based methods using our da ![Generated image examples](https://developer-blogs.nvidia.com/wp-content/uploads/2024/06/generated-medical-image-method-comparison-1.png) -**Figure 1.** Qualitative comparison of generated images between baseline methods
(retrained using our large-scale dataset) and our method. +**Figure 1.** Qualitative comparison of generated images between baseline methods
(retrained using our large-scale dataset) and our method. The MAISI here refers to `maisi3d-ddpm`. @@ -58,21 +70,20 @@ We retrained several state-of-the-art diffusion model-based methods using our da ## Time Cost and GPU Memory Usage ### Inference Time Cost and GPU Memory Usage -| `output_size` | latent size |`autoencoder_sliding_window_infer_size` | `autoencoder_tp_num_splits` | Peak Memory | DM Time | VAE Time | -|---------------|:--------------------------------------:|:--------------------------------------:|:---------------------------:|:-----------:|:-------:|:--------:| -| [256x256x128](./configs/config_infer_16g_256x256x128.json) |4x64x64x32| >=[64,64,32], not used | 2 | 14G | 57s | 1s | -| [256x256x256](./configs/config_infer_16g_256x256x256.json) |4x64x64x64| [48,48,64], 4 patches | 2 | 14G | 81s | 7s | -| [512x512x128](./configs/config_infer_16g_512x512x128.json) |4x128x128x32| [64,64,32], 9 patches | 1 | 14G | 138s | 7s | -| | | | | | | -| [256x256x256](./configs/config_infer_24g_256x256x256.json) |4x64x64x64| >=[64,64,64], not used | 4 | 22G | 81s | 2s | -| [512x512x128](./configs/config_infer_24g_512x512x128.json) |4x128x128x32| [80,80,32], 4 patches | 1 | 18G | 138s | 9s | -| [512x512x512](./configs/config_infer_24g_512x512x512.json) |4x128x128x128| [64,64,48], 36 patches | 2 | 22G | 569s | 29s | -| | | | | | | -| [512x512x512](./configs/config_infer_32g_512x512x512.json) |4x128x128x128| [64,64,64], 27 patches | 2 | 26G | 569s | 40s | -| | | | | | | -| [512x512x128](./configs/config_infer_80g_512x512x128.json) |4x128x128x32| >=[128,128,32], not used | 4 | 37G | 138s | 140s | -| [512x512x512](./configs/config_infer_80g_512x512x512.json) |4x128x128x128| [80,80,80], 8 patches | 2 | 44G | 569s | 30s | -| [512x512x768](./configs/config_infer_24g_512x512x768.json) |4x128x128x192| [80,80,112], 8 patches | 4 | 55G | 904s | 48s | +| `output_size` | Peak Memory | VAE Time + DM Time (`maisi3d-ddpm`) | VAE Time + DM Time (`maisi3d-rflow`) | latent size | `autoencoder_sliding_window_infer_size` | `autoencoder_tp_num_splits` | VAE Time | DM Time (`maisi3d-ddpm`) | DM Time (`maisi3d-rflow`) | +|---------------|:-----------:|:------------------------:|:------------------------:|:--------------------------------------:|:--------------------------------------:|:---------------------------:|:--------:|:---------------:|:---------------:| +| [256x256x128](./configs/config_infer_16g_256x256x128.json) | 15.0G | 58s | 3s | 4x64x64x32 | >=[64,64,32], not used | 2 | 1s | 57s | 2s | +| [256x256x256](./configs/config_infer_16g_256x256x256.json) | 15.4G | 86s | 8s | 4x64x64x64 | [48,48,64], 4 patches | 4 | 5s | 81s | 3s | +| [512x512x128](./configs/config_infer_16g_512x512x128.json) | 15.7G | 146s | 13s | 4x128x128x32 | [64,64,32], 9 patches | 2 | 8s | 138s | 5s | +| | | | | | | | | | | +| [256x256x256](./configs/config_infer_24g_256x256x256.json) | 22.7G | 83s | 5s | 4x64x64x64 | >=[64,64,64], not used | 4 | 2s | 81s | 3s | +| [512x512x128](./configs/config_infer_24g_512x512x128.json) | 21.0G | 144s | 11s | 4x128x128x32 | [80,80,32], 4 patches | 2 | 6s | 138s | 5s | +| [512x512x512](./configs/config_infer_24g_512x512x512.json) | 22.8G | 598s | 48s | 4x128x128x128 | [64,64,48], 36 patches | 2 | 29s | 569s | 19s | +| | | | | | | | | | | +| [512x512x512](./configs/config_infer_32g_512x512x512.json) | 28.4G | 599s | 49s | 4x128x128x128 | [80,80,48], 16 patches | 4 | 30s | 569s | 19s | +| | | | | | | | | | | +| [512x512x512](./configs/config_infer_80g_512x512x512.json) | 45.3G | 601s | 51s | 4x128x128x128 | [80,80,80], 8 patches | 2 | 32s | 569s | 19s | +| [512x512x768](./configs/config_infer_80g_512x512x768.json) | 49.7G | 961s | 87s | 4x128x128x192 | [80,80,96], 12 patches | 4 | 57s | 904s | 30s | **Table 3:** Inference Time Cost and GPU Memory Usage. `DM Time` refers to the time required for diffusion model inference. `VAE Time` refers to the time required for VAE decoder inference. The total inference time is the sum of `DM Time` and `VAE Time`. The experiment was conducted on an A100 80G GPU. @@ -86,7 +97,7 @@ When `autoencoder_sliding_window_infer_size` is equal to or larger than the late ### Training GPU Memory Usage The VAE is trained on patches and can be trained using a 16G GPU if the patch size is set to a small value, such as [64, 64, 64]. Users can adjust the patch size to fit the available GPU memory. For the released model, we initially trained the autoencoder on 16G V100 GPUs with a small patch size of [64, 64, 64], and then continued training on 32G V100 GPUs with a larger patch size of [128, 128, 128]. -The DM and ControlNet are trained on whole images rather than patches. The GPU memory usage during training depends on the size of the input images. +The DM and ControlNet are trained on whole images rather than patches. The GPU memory usage during training depends on the size of the input images. There is no big difference on memory usage between `maisi3d-ddpm` and `maisi3d-rflow`. | image size | latent size | Peak Memory | |--------------|:------------- |:-----------:| @@ -104,13 +115,13 @@ The DM and ControlNet are trained on whole images rather than patches. The GPU m ## MAISI Model Workflow The training and inference workflows of MAISI are depicted in the figure below. It begins by training an autoencoder in pixel space to encode images into latent features. Following that, it trains a diffusion model in the latent space to denoise the noisy latent features. During inference, it first generates latent features from random noise by applying multiple denoising steps using the trained diffusion model. Finally, it decodes the denoised latent features into images using the trained autoencoder.

- MAISI training scheme + MAISI training scheme
Figure 1: MAISI training scheme

- MAISI inference scheme Figure 2: MAISI inference scheme

@@ -120,6 +131,8 @@ MAISI is based on the following papers: [**ControlNet:** Lvmin Zhang, Anyi Rao, Maneesh Agrawala; β€œAdding Conditional Control to Text-to-Image Diffusion Models.” ICCV 2023.](https://openaccess.thecvf.com/content/ICCV2023/papers/Zhang_Adding_Conditional_Control_to_Text-to-Image_Diffusion_Models_ICCV_2023_paper.pdf) +[**Rectified Flow:** Liu, Xingchao, Chengyue Gong, and Qiang Liu. "Flow straight and fast: Learning to generate and transfer data with rectified flow." ICLR 2023](https://arxiv.org/pdf/2209.03003). + ### 1. Network Definition Network definition is stored in [./configs/config_maisi.json](./configs/config_maisi.json). Training and inference should use the same [./configs/config_maisi.json](./configs/config_maisi.json). @@ -131,7 +144,7 @@ The information for the inference input, such as the body region and anatomy to - `"spacing"`: The voxel size of the generated images. For example, if set to `[1.5, 1.5, 2.0]`, it generates images with a resolution of 1.5x1.5x2.0 mm. - `"output_size"`: The volume size of the generated images. For example, if set to `[512, 512, 256]`, it generates images of size 512x512x256. The values must be divisible by 16. If GPU memory is limited, adjust these to smaller numbers. Note that `"spacing"` and `"output_size"` together determine the output field of view (FOV). For example, if set to `[1.5, 1.5, 2.0]` mm and `[512, 512, 256]`, the FOV is 768x768x512 mm. We recommend the FOV in the x and y axes to be at least 256 mm for the head and at least 384 mm for other body regions like the abdomen. There is no restriction for the z-axis. - `"controllable_anatomy_size"`: A list specifying controllable anatomy and their size scale (0–1). For example, if set to `[["liver", 0.5], ["hepatic tumor", 0.3]]`, the generated image will contain a liver of median size (around the 50th percentile) and a relatively small hepatic tumor (around the 30th percentile). The output will include paired images and segmentation masks for the controllable anatomy. -- `"body_region"`: If `"controllable_anatomy_size"` is not specified, `"body_region"` will constrain the region of the generated images. It must be chosen from `"head"`, `"chest"`, `"thorax"`, `"abdomen"`, `"pelvis"`, or `"lower"`. Please set a reasonable `"body_region"` for the given FOV determined by `"spacing"` and `"output_size"`. For example, if FOV is only 128mm in z-axis, we should not expect `"body_region"` to contain all of [`"head"`, `"chest"`, `"thorax"`, `"abdomen"`, `"pelvis"`, `"lower"`]. +- `"body_region"`: For `maisi3d_rflow`, it is deprecated and can be set as `[]`. The output body region will be determined by `"anatomy_list"`. For `maisi3d_ddpm`, if `"controllable_anatomy_size"` is not specified, `"body_region"` will constrain the region of the generated images. It must be chosen from `"head"`, `"chest"`, `"thorax"`, `"abdomen"`, `"pelvis"`, or `"lower"`. Please set a reasonable `"body_region"` for the given FOV determined by `"spacing"` and `"output_size"`. For example, if FOV is only 128mm in z-axis, we should not expect `"body_region"` to contain all of [`"head"`, `"chest"`, `"thorax"`, `"abdomen"`, `"pelvis"`, `"lower"`]. - `"anatomy_list"`: If `"controllable_anatomy_size"` is not specified, the output will include paired images and segmentation masks for the anatomy listed in `"./configs/label_dict.json"`. - `"autoencoder_sliding_window_infer_size"`: To save GPU memory, sliding window inference is used when decoding latents into images if `"output_size"` is large. This parameter specifies the patch size of the sliding window. Smaller values reduce GPU memory usage but increase the time cost. The values must be divisible by 16. If GPU memory is sufficient, select a larger value for this parameter. - `"autoencoder_sliding_window_infer_overlap"`: A float between 0 and 1. Larger values reduce stitching artifacts when patches are stitched during sliding window inference but increase the time cost. If you do not observe seam lines in the generated image, you can use a smaller value to save inference time. @@ -164,12 +177,21 @@ For example, |[512, 512, 512] | [1.0, 1.0, 1.0] | #### Execute Inference: -To run the inference script, please run: +To run the inference script with MAISI DDPM, please set `"num_inference_steps": 1000` in `./configs/config_infer.json`, and run: ```bash export MONAI_DATA_DIRECTORY= -python -m scripts.inference -c ./configs/config_maisi.json -i ./configs/config_infer.json -e ./configs/environment.json --random-seed 0 +python -m scripts.inference -c ./configs/config_maisi3d-ddpm.json -i ./configs/config_infer.json -e ./configs/environment_maisi3d-ddpm.json --random-seed 0 --version maisi3d-ddpm ``` +To run the inference script with MAISI RFlow, please set `"num_inference_steps": 30` in `./configs/config_infer.json`, and run: +```bash +export MONAI_DATA_DIRECTORY= +python -m scripts.inference -c ./configs/config_maisi3d-rflow.json -i ./configs/config_infer.json -e ./configs/environment_maisi3d-rflow.json --random-seed 0 --version maisi3d-rflow +``` + +If GPU OOM happens, please increase `autoencoder_tp_num_splits` or reduce `autoencoder_sliding_window_infer_size` in `./configs/config_infer.json`. +To reduce time cost, please reduce `autoencoder_sliding_window_infer_overlap` in `./configs/config_infer.json`, while monitoring whether stitching artifact occurs. + Please refer to [maisi_inference_tutorial.ipynb](maisi_inference_tutorial.ipynb) for the tutorial for MAISI model inference. @@ -177,7 +199,12 @@ Please refer to [maisi_inference_tutorial.ipynb](maisi_inference_tutorial.ipynb) To run the inference script with TensorRT acceleration, please run: ```bash export MONAI_DATA_DIRECTORY= -python -m scripts.inference -c ./configs/config_maisi.json -i ./configs/config_infer.json -e ./configs/environment.json -x ./configs/config_trt.json --random-seed 0 +python -m scripts.inference -c ./configs/config_maisi3d-ddpm.json -i ./configs/config_infer.json -e ./configs/environment_maisi3d-ddpm.json -x ./configs/config_trt.json --random-seed 0 --version maisi3d-ddpm +``` + +```bash +export MONAI_DATA_DIRECTORY= +python -m scripts.inference -c ./configs/config_maisi3d-rflow.json -i ./configs/config_infer.json -e ./configs/environment_maisi3d-rflow.json -x ./configs/config_trt.json --random-seed 0 --version maisi3d-rflow ``` Extra config file, [./configs/config_trt.json](./configs/config_trt.json) is using `trt_compile()` utility from MONAI to convert select modules to TensorRT by overriding their definitions from [./configs/config_infer.json](./configs/config_infer.json). @@ -215,7 +242,7 @@ Please refer to [maisi_train_vae_tutorial.ipynb](maisi_train_vae_tutorial.ipynb) #### [3.2 3D Latent Diffusion Training](./scripts/diff_model_train.py) -Please refer to [maisi_diff_unet_training_tutorial.ipynb](maisi_diff_unet_training_tutorial.ipynb) for the tutorial for MAISI diffusion model training. +Please refer to [maisi_train_diff_unet_tutorial.ipynb](maisi_train_diff_unet_tutorial.ipynb) for the tutorial for MAISI diffusion model training. #### [3.3 3D ControlNet Training](./scripts/train_controlnet.py) @@ -233,7 +260,11 @@ The training was performed with the following: #### Execute Training: To train with a single GPU, please run: ```bash -python -m scripts.train_controlnet -c ./configs/config_maisi.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g 1 +python -m scripts.train_controlnet -c ./configs/config_maisi3d-ddpm.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g 1 +``` + +```bash +python -m scripts.train_controlnet -c ./configs/config_maisi3d-rflow.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g 1 ``` The training script also enables multi-GPU training. For instance, if you are using eight GPUs, you can run the training script with the following command: @@ -243,7 +274,16 @@ torchrun \ --nproc_per_node=${NUM_GPUS_PER_NODE} \ --nnodes=1 \ --master_addr=localhost --master_port=1234 \ - -m scripts.train_controlnet -c ./configs/config_maisi.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g ${NUM_GPUS_PER_NODE} + -m scripts.train_controlnet -c ./configs/config_maisi3d-ddpm.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g ${NUM_GPUS_PER_NODE} +``` + +```bash +export NUM_GPUS_PER_NODE=8 +torchrun \ + --nproc_per_node=${NUM_GPUS_PER_NODE} \ + --nnodes=1 \ + --master_addr=localhost --master_port=1234 \ + -m scripts.train_controlnet -c ./configs/config_maisi3d-rflow.json -t ./configs/config_maisi_controlnet_train.json -e ./configs/environment_maisi_controlnet_train.json -g ${NUM_GPUS_PER_NODE} ``` Please also check [maisi_train_controlnet_tutorial.ipynb](./maisi_train_controlnet_tutorial.ipynb) for more details about data preparation and training parameters. diff --git a/generation/maisi/configs/config_infer.json b/generation/maisi/configs/config_infer.json index fc08a7bda4..19b12a6d22 100644 --- a/generation/maisi/configs/config_infer.json +++ b/generation/maisi/configs/config_infer.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 256, @@ -18,10 +18,11 @@ 2.0 ], "autoencoder_sliding_window_infer_size": [48,48,48], - "autoencoder_sliding_window_infer_overlap": 0.25, + "autoencoder_sliding_window_infer_overlap": 0.6666, "controlnet": "$@controlnet_def", "diffusion_unet": "$@diffusion_unet_def", "autoencoder": "$@autoencoder_def", "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", - "mask_generation_diffusion": "$@mask_generation_diffusion_def" + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_16g_256x256x128.json b/generation/maisi/configs/config_infer_16g_256x256x128.json index 72933304ba..1c6d424f2e 100644 --- a/generation/maisi/configs/config_infer_16g_256x256x128.json +++ b/generation/maisi/configs/config_infer_16g_256x256x128.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 256, @@ -19,5 +19,11 @@ ], "autoencoder_sliding_window_infer_size": [96,96,96], "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 2 + "autoencoder_tp_num_splits": 2, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_16g_256x256x256.json b/generation/maisi/configs/config_infer_16g_256x256x256.json index d4ec9e1a88..8ccd0bc2ca 100644 --- a/generation/maisi/configs/config_infer_16g_256x256x256.json +++ b/generation/maisi/configs/config_infer_16g_256x256x256.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 256, @@ -18,6 +18,12 @@ 2.0 ], "autoencoder_sliding_window_infer_size": [48,48,64], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 2 + "autoencoder_sliding_window_infer_overlap": 0.6666, + "autoencoder_tp_num_splits": 4, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_16g_512x512x128.json b/generation/maisi/configs/config_infer_16g_512x512x128.json index 5e067cd4b4..ec80d72a84 100644 --- a/generation/maisi/configs/config_infer_16g_512x512x128.json +++ b/generation/maisi/configs/config_infer_16g_512x512x128.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -18,6 +18,12 @@ 4.0 ], "autoencoder_sliding_window_infer_size": [64,64,32], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 1 + "autoencoder_sliding_window_infer_overlap": 0.5, + "autoencoder_tp_num_splits": 2, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_24g_256x256x256.json b/generation/maisi/configs/config_infer_24g_256x256x256.json index bb0806f635..a0be706b1c 100644 --- a/generation/maisi/configs/config_infer_24g_256x256x256.json +++ b/generation/maisi/configs/config_infer_24g_256x256x256.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 256, @@ -19,5 +19,11 @@ ], "autoencoder_sliding_window_infer_size": [64,64,64], "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 4 + "autoencoder_tp_num_splits": 4, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_24g_512x512x128.json b/generation/maisi/configs/config_infer_24g_512x512x128.json index 6d2b9d7eab..95bd38795a 100644 --- a/generation/maisi/configs/config_infer_24g_512x512x128.json +++ b/generation/maisi/configs/config_infer_24g_512x512x128.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -18,6 +18,12 @@ 4.0 ], "autoencoder_sliding_window_infer_size": [80,80,32], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 1 + "autoencoder_sliding_window_infer_overlap": 0.4, + "autoencoder_tp_num_splits": 2, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_24g_512x512x512.json b/generation/maisi/configs/config_infer_24g_512x512x512.json index 2cbfb9573f..0606c5e945 100644 --- a/generation/maisi/configs/config_infer_24g_512x512x512.json +++ b/generation/maisi/configs/config_infer_24g_512x512x512.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -18,6 +18,12 @@ 1.0 ], "autoencoder_sliding_window_infer_size": [64,64,48], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 2 + "autoencoder_sliding_window_infer_overlap": 0.4, + "autoencoder_tp_num_splits": 2, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_32g_512x512x512.json b/generation/maisi/configs/config_infer_32g_512x512x512.json index 5dcbcacbe0..5044955c99 100644 --- a/generation/maisi/configs/config_infer_32g_512x512x512.json +++ b/generation/maisi/configs/config_infer_32g_512x512x512.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -17,7 +17,13 @@ 0.75, 1.0 ], - "autoencoder_sliding_window_infer_size": [64,64,64], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 2 + "autoencoder_sliding_window_infer_size": [80,80,48], + "autoencoder_sliding_window_infer_overlap": 0.4, + "autoencoder_tp_num_splits": 4, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_80g_512x512x128.json b/generation/maisi/configs/config_infer_80g_512x512x128.json deleted file mode 100644 index d20dbbc76b..0000000000 --- a/generation/maisi/configs/config_infer_80g_512x512x128.json +++ /dev/null @@ -1,23 +0,0 @@ -{ - "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], - "controllable_anatomy_size": [], - "num_inference_steps": 1000, - "mask_generation_num_inference_steps": 1000, - "output_size": [ - 512, - 512, - 128 - ], - "image_output_ext": ".nii.gz", - "label_output_ext": ".nii.gz", - "spacing": [ - 0.75, - 0.75, - 4.0 - ], - "autoencoder_sliding_window_infer_size": [128,128,32], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 4 -} diff --git a/generation/maisi/configs/config_infer_80g_512x512x512.json b/generation/maisi/configs/config_infer_80g_512x512x512.json index bfcd6b7dc7..71f5d031e3 100644 --- a/generation/maisi/configs/config_infer_80g_512x512x512.json +++ b/generation/maisi/configs/config_infer_80g_512x512x512.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, - "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "body_region": ["chest"], + "anatomy_list": ["lung tumor"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -18,6 +18,12 @@ 1.0 ], "autoencoder_sliding_window_infer_size": [80,80,80], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 2 + "autoencoder_sliding_window_infer_overlap": 0.4, + "autoencoder_tp_num_splits": 2, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_infer_80g_512x512x768.json b/generation/maisi/configs/config_infer_80g_512x512x768.json index 9cb7e61b61..9d1bee4cd2 100644 --- a/generation/maisi/configs/config_infer_80g_512x512x768.json +++ b/generation/maisi/configs/config_infer_80g_512x512x768.json @@ -1,9 +1,9 @@ { "num_output_samples": 1, "body_region": ["abdomen"], - "anatomy_list": ["liver","hepatic tumor"], + "anatomy_list": ["liver"], "controllable_anatomy_size": [], - "num_inference_steps": 1000, + "num_inference_steps": 30, "mask_generation_num_inference_steps": 1000, "output_size": [ 512, @@ -17,7 +17,13 @@ 0.75, 0.66667 ], - "autoencoder_sliding_window_infer_size": [80,80,112], - "autoencoder_sliding_window_infer_overlap": 0.25, - "autoencoder_tp_num_splits": 4 + "autoencoder_sliding_window_infer_size": [80,80,96], + "autoencoder_sliding_window_infer_overlap": 0.4, + "autoencoder_tp_num_splits": 4, + "controlnet": "$@controlnet_def", + "diffusion_unet": "$@diffusion_unet_def", + "autoencoder": "$@autoencoder_def", + "mask_generation_autoencoder": "$@mask_generation_autoencoder_def", + "mask_generation_diffusion": "$@mask_generation_diffusion_def", + "modality": 1 } diff --git a/generation/maisi/configs/config_maisi.json b/generation/maisi/configs/config_maisi3d-ddpm.json similarity index 96% rename from generation/maisi/configs/config_maisi.json rename to generation/maisi/configs/config_maisi3d-ddpm.json index 8a781ca5b4..67b38e9c15 100644 --- a/generation/maisi/configs/config_maisi.json +++ b/generation/maisi/configs/config_maisi3d-ddpm.json @@ -2,6 +2,7 @@ "spatial_dims": 3, "image_channels": 1, "latent_channels": 4, + "include_body_region": true, "mask_generation_latent_shape": [ 4, 64, @@ -60,8 +61,8 @@ ], "num_res_blocks": 2, "use_flash_attention": true, - "include_top_region_index_input": true, - "include_bottom_region_index_input": true, + "include_top_region_index_input": "@include_body_region", + "include_bottom_region_index_input": "@include_body_region", "include_spacing_input": true }, "controlnet_def": { diff --git a/generation/maisi/configs/config_maisi3d-rflow.json b/generation/maisi/configs/config_maisi3d-rflow.json new file mode 100644 index 0000000000..d76da08bf3 --- /dev/null +++ b/generation/maisi/configs/config_maisi3d-rflow.json @@ -0,0 +1,150 @@ +{ + "spatial_dims": 3, + "image_channels": 1, + "latent_channels": 4, + "include_body_region": false, + "mask_generation_latent_shape": [ + 4, + 64, + 64, + 64 + ], + "autoencoder_def": { + "_target_": "monai.apps.generation.maisi.networks.autoencoderkl_maisi.AutoencoderKlMaisi", + "spatial_dims": "@spatial_dims", + "in_channels": "@image_channels", + "out_channels": "@image_channels", + "latent_channels": "@latent_channels", + "num_channels": [ + 64, + 128, + 256 + ], + "num_res_blocks": [2,2,2], + "norm_num_groups": 32, + "norm_eps": 1e-06, + "attention_levels": [ + false, + false, + false + ], + "with_encoder_nonlocal_attn": false, + "with_decoder_nonlocal_attn": false, + "use_checkpointing": false, + "use_convtranspose": false, + "norm_float16": true, + "num_splits": 4, + "dim_split": 1 + }, + "diffusion_unet_def": { + "_target_": "monai.apps.generation.maisi.networks.diffusion_model_unet_maisi.DiffusionModelUNetMaisi", + "spatial_dims": "@spatial_dims", + "in_channels": "@latent_channels", + "out_channels": "@latent_channels", + "num_channels": [64, 128, 256, 512], + "attention_levels": [ + false, + false, + true, + true + ], + "num_head_channels": [ + 0, + 0, + 32, + 32 + ], + "num_res_blocks": 2, + "use_flash_attention": true, + "include_top_region_index_input": "@include_body_region", + "include_bottom_region_index_input": "@include_body_region", + "include_spacing_input": true, + "num_class_embeds": 128, + "resblock_updown": true, + "include_fc": true + }, + "controlnet_def": { + "_target_": "monai.apps.generation.maisi.networks.controlnet_maisi.ControlNetMaisi", + "spatial_dims": "@spatial_dims", + "in_channels": "@latent_channels", + "num_channels": [64, 128, 256, 512], + "attention_levels": [ + false, + false, + true, + true + ], + "num_head_channels": [ + 0, + 0, + 32, + 32 + ], + "num_res_blocks": 2, + "use_flash_attention": true, + "conditioning_embedding_in_channels": 8, + "conditioning_embedding_num_channels": [8, 32, 64], + "num_class_embeds": 128, + "resblock_updown": true, + "include_fc": true + }, + "mask_generation_autoencoder_def": { + "_target_": "monai.apps.generation.maisi.networks.autoencoderkl_maisi.AutoencoderKlMaisi", + "spatial_dims": "@spatial_dims", + "in_channels": 8, + "out_channels": 125, + "latent_channels": "@latent_channels", + "num_channels": [ + 32, + 64, + 128 + ], + "num_res_blocks": [1, 2, 2], + "norm_num_groups": 32, + "norm_eps": 1e-06, + "attention_levels": [ + false, + false, + false + ], + "with_encoder_nonlocal_attn": false, + "with_decoder_nonlocal_attn": false, + "use_flash_attention": false, + "use_checkpointing": true, + "use_convtranspose": true, + "norm_float16": true, + "num_splits": 8, + "dim_split": 1 + }, + "mask_generation_diffusion_def": { + "_target_": "monai.networks.nets.diffusion_model_unet.DiffusionModelUNet", + "spatial_dims": "@spatial_dims", + "in_channels": "@latent_channels", + "out_channels": "@latent_channels", + "channels":[64, 128, 256, 512], + "attention_levels":[false, false, true, true], + "num_head_channels":[0, 0, 32, 32], + "num_res_blocks": 2, + "use_flash_attention": true, + "with_conditioning": true, + "upcast_attention": true, + "cross_attention_dim": 10 + }, + "mask_generation_scale_factor": 1.0055984258651733, + "noise_scheduler": { + "_target_": "monai.networks.schedulers.rectified_flow.RFlowScheduler", + "num_train_timesteps": 1000, + "use_discrete_timesteps": false, + "use_timestep_transform": true, + "sample_method": "uniform", + "scale":1.4 + }, + "mask_generation_noise_scheduler": { + "_target_": "monai.networks.schedulers.ddpm.DDPMScheduler", + "num_train_timesteps": 1000, + "beta_start": 0.0015, + "beta_end": 0.0195, + "schedule": "scaled_linear_beta", + "clip_sample": false + } +} diff --git a/generation/maisi/configs/config_maisi_controlnet_train.json b/generation/maisi/configs/config_maisi_controlnet_train.json index 4ac94efe63..9bca444bd4 100644 --- a/generation/maisi/configs/config_maisi_controlnet_train.json +++ b/generation/maisi/configs/config_maisi_controlnet_train.json @@ -9,7 +9,9 @@ "weighted_loss": 100 }, "controlnet_infer": { - "num_inference_steps": 1000, - "autoencoder_sliding_window_infer_size": [96, 96, 96] + "num_inference_steps": 10, + "autoencoder_sliding_window_infer_size": [80, 80, 80], + "autoencoder_sliding_window_infer_overlap": 0.4, + "modality": 1 } } diff --git a/generation/maisi/configs/config_maisi_diff_model.json b/generation/maisi/configs/config_maisi_diff_model.json index 8407dbdcc1..f97a749c89 100644 --- a/generation/maisi/configs/config_maisi_diff_model.json +++ b/generation/maisi/configs/config_maisi_diff_model.json @@ -29,6 +29,7 @@ 0 ], "random_seed": 0, - "num_inference_steps": 10 + "num_inference_steps": 10, + "modality": 1 } } diff --git a/generation/maisi/configs/environment.json b/generation/maisi/configs/environment_maisi3d-ddpm.json similarity index 75% rename from generation/maisi/configs/environment.json rename to generation/maisi/configs/environment_maisi3d-ddpm.json index 5e017645f1..97ba2b590a 100644 --- a/generation/maisi/configs/environment.json +++ b/generation/maisi/configs/environment_maisi3d-ddpm.json @@ -1,8 +1,8 @@ { "output_dir": "output", "trained_autoencoder_path": "models/autoencoder_epoch273.pt", - "trained_diffusion_path": "models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt", - "trained_controlnet_path": "models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt", + "trained_diffusion_path": "models/diff_unet_3d_ddpm.pt", + "trained_controlnet_path": "models/controlnet_3d_ddpm.pt", "trained_mask_generation_autoencoder_path": "models/mask_generation_autoencoder.pt", "trained_mask_generation_diffusion_path": "models/mask_generation_diffusion_unet.pt", "all_mask_files_base_dir": "datasets/all_masks_flexible_size_and_spacing_3000", diff --git a/generation/maisi/configs/environment_maisi3d-rflow.json b/generation/maisi/configs/environment_maisi3d-rflow.json new file mode 100644 index 0000000000..c1be37b592 --- /dev/null +++ b/generation/maisi/configs/environment_maisi3d-rflow.json @@ -0,0 +1,13 @@ +{ + "output_dir": "output", + "trained_autoencoder_path": "models/autoencoder_epoch273.pt", + "trained_diffusion_path": "models/diff_unet_3d_rflow.pt", + "trained_controlnet_path": "models/controlnet_3d_rflow.pt", + "trained_mask_generation_autoencoder_path": "models/mask_generation_autoencoder.pt", + "trained_mask_generation_diffusion_path": "models/mask_generation_diffusion_unet.pt", + "all_mask_files_base_dir": "datasets/all_masks_flexible_size_and_spacing_4000", + "all_mask_files_json": "./configs/candidate_masks_flexible_size_and_spacing_4000.json", + "all_anatomy_size_conditions_json": "./configs/all_anatomy_size_condtions.json", + "label_dict_json": "./configs/label_dict.json", + "label_dict_remap_json": "./configs/label_dict_124_to_132.json" +} diff --git a/generation/maisi/configs/modality_mapping.json b/generation/maisi/configs/modality_mapping.json new file mode 100644 index 0000000000..38bd3ee321 --- /dev/null +++ b/generation/maisi/configs/modality_mapping.json @@ -0,0 +1,15 @@ +{ + "unknown":0, + "ct":1, + "ct_wo_contrast":2, + "ct_contrast":3, + "mri":8, + "mri_t1":9, + "mri_t2":10, + "mri_flair":11, + "mri_pd":12, + "mri_dwi":13, + "mri_adc":14, + "mri_ssfp":15, + "mri_mra":16 +} diff --git a/generation/maisi/figures/maisi_infer.jpg b/generation/maisi/figures/maisi_infer.jpg deleted file mode 100644 index 9210da5fdd..0000000000 Binary files a/generation/maisi/figures/maisi_infer.jpg and /dev/null differ diff --git a/generation/maisi/figures/maisi_infer.png b/generation/maisi/figures/maisi_infer.png new file mode 100644 index 0000000000..4bd18ea188 Binary files /dev/null and b/generation/maisi/figures/maisi_infer.png differ diff --git a/generation/maisi/figures/maisi_train.jpg b/generation/maisi/figures/maisi_train.jpg deleted file mode 100644 index 8c4936456d..0000000000 Binary files a/generation/maisi/figures/maisi_train.jpg and /dev/null differ diff --git a/generation/maisi/figures/maisi_train.png b/generation/maisi/figures/maisi_train.png new file mode 100644 index 0000000000..d0ec7fda8e Binary files /dev/null and b/generation/maisi/figures/maisi_train.png differ diff --git a/generation/maisi/maisi_inference_tutorial.ipynb b/generation/maisi/maisi_inference_tutorial.ipynb index d121f886b5..69ce91b9b5 100644 --- a/generation/maisi/maisi_inference_tutorial.ipynb +++ b/generation/maisi/maisi_inference_tutorial.ipynb @@ -18,7 +18,9 @@ "\n", "# MAISI Inference Tutorial\n", "\n", - "This tutorial illustrates how to use trained MAISI model and codebase to generate synthetic 3D images and paired masks." + "This tutorial illustrates how to use trained MAISI model and codebase to generate synthetic 3D images and paired masks.\n", + "\n", + "`[Release Note (March 2025)]:` We are excited to announce the new MAISI Version `'maisi3d-rflow'`. Compared with the previous version `'maisi3d-ddpm'`, it accelerated latent diffusion model inference by 33x. Please see the detailed difference in the following section." ] }, { @@ -61,32 +63,32 @@ "name": "stdout", "output_type": "stream", "text": [ - "MONAI version: 1.4.0rc10\n", - "Numpy version: 1.24.4\n", - "Pytorch version: 2.5.0a0+872d972e41.nv24.08.01\n", + "MONAI version: 1.4.1rc1+32.g34f37973\n", + "Numpy version: 1.26.4\n", + "Pytorch version: 2.5.0+cu124\n", "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n", - "MONAI rev id: cac21f6936a2e8d6e4e57e4e958f8e32aae1585e\n", - "MONAI __file__: /usr/local/lib/python3.10/dist-packages/monai/__init__.py\n", + "MONAI rev id: 34f379735c5e18e7f809453eb1b3606c225c788b\n", + "MONAI __file__: /localhome//.local/lib/python3.10/site-packages/monai/__init__.py\n", "\n", "Optional dependencies:\n", "Pytorch Ignite version: 0.4.11\n", "ITK version: 5.4.0\n", - "Nibabel version: 5.2.1\n", - "scikit-image version: 0.23.2\n", - "scipy version: 1.13.1\n", - "Pillow version: 10.4.0\n", - "Tensorboard version: 2.17.0\n", + "Nibabel version: 5.3.2\n", + "scikit-image version: 0.24.0\n", + "scipy version: 1.14.1\n", + "Pillow version: 11.0.0\n", + "Tensorboard version: 2.18.0\n", "gdown version: 5.2.0\n", - "TorchVision version: 0.20.0a0\n", - "tqdm version: 4.66.4\n", + "TorchVision version: 0.20.0+cu124\n", + "tqdm version: 4.66.5\n", "lmdb version: 1.5.1\n", - "psutil version: 5.9.8\n", - "pandas version: 2.2.2\n", - "einops version: 0.7.0\n", + "psutil version: 6.1.0\n", + "pandas version: 2.2.3\n", + "einops version: 0.8.0\n", "transformers version: 4.40.2\n", - "mlflow version: 2.16.0\n", + "mlflow version: 2.17.1\n", "pynrrd version: 1.0.0\n", - "clearml version: 1.16.3\n", + "clearml version: 1.16.5rc2\n", "\n", "For details about installing the optional dependencies, please visit:\n", " https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n", @@ -109,8 +111,52 @@ "from scripts.sample import LDMSampler, check_input\n", "from scripts.utils import define_instance\n", "from scripts.utils_plot import find_label_center_loc, get_xyz_plot, show_image\n", + "from scripts.diff_model_setting import setup_logging\n", + "\n", + "print_config()\n", + "\n", + "logger = setup_logging(\"notebook\")" + ] + }, + { + "cell_type": "markdown", + "id": "37c2759e-d1fa-42d7-b208-fbe306ac1e06", + "metadata": {}, + "source": [ + "## Set up the MAISI version\n", "\n", - "print_config()" + "Choose between `'maisi3d-ddpm'` and `'maisi3d-rflow'`. The differences are:\n", + "- The maisi version `'maisi3d-ddpm'` uses basic noise scheduler DDPM. `'maisi3d-rflow'` uses Rectified Flow scheduler, can be 33 times faster during inference.\n", + "- The maisi version `'maisi3d-ddpm'` requires training images to be labeled with body region (`\"top_region_index\"` and `\"bottom_region_index\"`), while `'maisi3d-rflow'` does not have such requirement. In other words, it is easier to prepare training data for `'maisi3d-rflow'`.\n", + "- For the released model weights, `'maisi3d-rflow'` can generate images with better quality for head region and small output volumes, and comparable quality for other cases compared with `'maisi3d-ddpm'`." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "bf4252b1-089d-48c1-b6d6-aa24a93a5839", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "[2025-03-19 05:06:22.467][ INFO](notebook) - MAISI version is maisi3d-rflow, whether to use body_region is False\n" + ] + } + ], + "source": [ + "maisi_version = \"maisi3d-rflow\"\n", + "if maisi_version == \"maisi3d-ddpm\":\n", + " model_def_path = \"./configs/config_maisi3d-ddpm.json\"\n", + "elif maisi_version == \"maisi3d-rflow\":\n", + " model_def_path = \"./configs/config_maisi3d-rflow.json\"\n", + "else:\n", + " raise ValueError(f\"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}.\")\n", + "with open(model_def_path, \"r\") as f:\n", + " model_def = json.load(f)\n", + "include_body_region = model_def[\"include_body_region\"]\n", + "logger.info(f\"MAISI version is {maisi_version}, whether to use body_region is {include_body_region}\")" ] }, { @@ -119,14 +165,12 @@ "metadata": {}, "source": [ "## Setup data directory\n", - "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable.\n", - "This allows you to save results and reuse downloads.\n", - "If not specified a temporary directory will be used." + "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable. This allows you to save results and reuse downloads. If not specified a temporary directory will be used." ] }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, "id": "e3c12dcc", "metadata": {}, "outputs": [ @@ -134,26 +178,27 @@ "name": "stdout", "output_type": "stream", "text": [ - "2024-09-30 06:40:49,932 - INFO - Expected md5 is None, skip md5 check for file models/autoencoder_epoch273.pt.\n", - "2024-09-30 06:40:49,933 - INFO - File exists: models/autoencoder_epoch273.pt, skipped downloading.\n", - "2024-09-30 06:40:49,933 - INFO - Expected md5 is None, skip md5 check for file models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt.\n", - "2024-09-30 06:40:49,933 - INFO - File exists: models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt, skipped downloading.\n", - "2024-09-30 06:40:49,934 - INFO - Expected md5 is None, skip md5 check for file models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt.\n", - "2024-09-30 06:40:49,934 - INFO - File exists: models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt, skipped downloading.\n", - "2024-09-30 06:40:49,934 - INFO - Expected md5 is None, skip md5 check for file models/mask_generation_autoencoder.pt.\n", - "2024-09-30 06:40:49,934 - INFO - File exists: models/mask_generation_autoencoder.pt, skipped downloading.\n", - "2024-09-30 06:40:49,935 - INFO - Expected md5 is None, skip md5 check for file models/mask_generation_diffusion_unet.pt.\n", - "2024-09-30 06:40:49,935 - INFO - File exists: models/mask_generation_diffusion_unet.pt, skipped downloading.\n", - "2024-09-30 06:40:49,935 - INFO - Expected md5 is None, skip md5 check for file configs/candidate_masks_flexible_size_and_spacing_3000.json.\n", - "2024-09-30 06:40:49,935 - INFO - File exists: configs/candidate_masks_flexible_size_and_spacing_3000.json, skipped downloading.\n", - "2024-09-30 06:40:49,936 - INFO - Expected md5 is None, skip md5 check for file configs/all_anatomy_size_condtions.json.\n", - "2024-09-30 06:40:49,936 - INFO - File exists: configs/all_anatomy_size_condtions.json, skipped downloading.\n", - "2024-09-30 06:40:49,936 - INFO - Expected md5 is None, skip md5 check for file /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000.zip.\n", - "2024-09-30 06:40:49,936 - INFO - File exists: /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000.zip, skipped downloading.\n" + "2025-03-19 05:06:22,476 - INFO - Expected md5 is None, skip md5 check for file models/autoencoder_epoch273.pt.\n", + "2025-03-19 05:06:22,477 - INFO - File exists: models/autoencoder_epoch273.pt, skipped downloading.\n", + "2025-03-19 05:06:22,478 - INFO - Expected md5 is None, skip md5 check for file models/mask_generation_autoencoder.pt.\n", + "2025-03-19 05:06:22,478 - INFO - File exists: models/mask_generation_autoencoder.pt, skipped downloading.\n", + "2025-03-19 05:06:22,479 - INFO - Expected md5 is None, skip md5 check for file models/mask_generation_diffusion_unet.pt.\n", + "2025-03-19 05:06:22,480 - INFO - File exists: models/mask_generation_diffusion_unet.pt, skipped downloading.\n", + "2025-03-19 05:06:22,481 - INFO - Expected md5 is None, skip md5 check for file configs/all_anatomy_size_condtions.json.\n", + "2025-03-19 05:06:22,482 - INFO - File exists: configs/all_anatomy_size_condtions.json, skipped downloading.\n", + "2025-03-19 05:06:22,482 - INFO - Expected md5 is None, skip md5 check for file temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000.zip.\n", + "2025-03-19 05:06:22,483 - INFO - File exists: temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000.zip, skipped downloading.\n", + "2025-03-19 05:06:22,483 - INFO - Expected md5 is None, skip md5 check for file models/diff_unet_3d_rflow.pt.\n", + "2025-03-19 05:06:22,484 - INFO - File exists: models/diff_unet_3d_rflow.pt, skipped downloading.\n", + "2025-03-19 05:06:22,484 - INFO - Expected md5 is None, skip md5 check for file models/controlnet_3d_rflow.pt.\n", + "2025-03-19 05:06:22,485 - INFO - File exists: models/controlnet_3d_rflow.pt, skipped downloading.\n", + "2025-03-19 05:06:22,485 - INFO - Expected md5 is None, skip md5 check for file configs/candidate_masks_flexible_size_and_spacing_4000.json.\n", + "2025-03-19 05:06:22,486 - INFO - File exists: configs/candidate_masks_flexible_size_and_spacing_4000.json, skipped downloading.\n" ] } ], "source": [ + "os.environ[\"MONAI_DATA_DIRECTORY\"] = \"temp_work_dir_inference_demo\"\n", "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n", "if directory is not None:\n", " os.makedirs(directory, exist_ok=True)\n", @@ -167,16 +212,6 @@ " \"/model_zoo/model_maisi_autoencoder_epoch273_alternative.pt\",\n", " },\n", " {\n", - " \"path\": \"models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt\",\n", - " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo\"\n", - " \"/model_maisi_input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1_alternative.pt\",\n", - " },\n", - " {\n", - " \"path\": \"models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt\",\n", - " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo\"\n", - " \"/model_maisi_controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current_alternative.pt\",\n", - " },\n", - " {\n", " \"path\": \"models/mask_generation_autoencoder.pt\",\n", " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai\" \"/tutorials/mask_generation_autoencoder.pt\",\n", " },\n", @@ -186,21 +221,54 @@ " \"/tutorials/model_zoo/model_maisi_mask_generation_diffusion_unet_v2.pt\",\n", " },\n", " {\n", - " \"path\": \"configs/candidate_masks_flexible_size_and_spacing_3000.json\",\n", - " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai\"\n", - " \"/tutorials/candidate_masks_flexible_size_and_spacing_3000.json\",\n", - " },\n", - " {\n", " \"path\": \"configs/all_anatomy_size_condtions.json\",\n", " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/all_anatomy_size_condtions.json\",\n", " },\n", " {\n", - " \"path\": \"datasets/all_masks_flexible_size_and_spacing_3000.zip\",\n", + " \"path\": \"datasets/all_masks_flexible_size_and_spacing_4000.zip\",\n", " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai\"\n", - " \"/tutorials/model_zoo/model_maisi_all_masks_flexible_size_and_spacing_3000.zip\",\n", + " \"/tutorials/all_masks_flexible_size_and_spacing_4000.zip\",\n", " },\n", "]\n", "\n", + "if maisi_version == \"maisi3d-ddpm\":\n", + " files += [\n", + " {\n", + " \"path\": \"models/diff_unet_3d_ddpm.pt\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo\"\n", + " \"/model_maisi_input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1_alternative.pt\",\n", + " },\n", + " {\n", + " \"path\": \"models/controlnet_3d_ddpm.pt\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo\"\n", + " \"/model_maisi_controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current_alternative.pt\",\n", + " },\n", + " {\n", + " \"path\": \"configs/candidate_masks_flexible_size_and_spacing_3000.json\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai\"\n", + " \"/tutorials/candidate_masks_flexible_size_and_spacing_3000.json\",\n", + " },\n", + " ]\n", + "elif maisi_version == \"maisi3d-rflow\":\n", + " files += [\n", + " {\n", + " \"path\": \"models/diff_unet_3d_rflow.pt\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/\"\n", + " \"diff_unet_ckpt_rflow_epoch19350.pt\",\n", + " },\n", + " {\n", + " \"path\": \"models/controlnet_3d_rflow.pt\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai/tutorials/controlnet_rflow_epoch60.pt\",\n", + " },\n", + " {\n", + " \"path\": \"configs/candidate_masks_flexible_size_and_spacing_4000.json\",\n", + " \"url\": \"https://developer.download.nvidia.com/assets/Clara/monai\"\n", + " \"/tutorials/candidate_masks_flexible_size_and_spacing_4000.json\",\n", + " },\n", + " ]\n", + "else:\n", + " raise ValueError(f\"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}.\")\n", + "\n", "for file in files:\n", " file[\"path\"] = file[\"path\"] if \"datasets/\" not in file[\"path\"] else os.path.join(root_dir, file[\"path\"])\n", " download_url(url=file[\"url\"], filepath=file[\"path\"])" @@ -218,43 +286,49 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 5, "id": "c38b4c33", "metadata": { "scrolled": true }, "outputs": [ { - "name": "stdout", + "name": "stderr", "output_type": "stream", "text": [ - "output_dir: output\n", - "trained_autoencoder_path: models/autoencoder_epoch273.pt\n", - "trained_diffusion_path: models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt\n", - "trained_controlnet_path: models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt\n", - "trained_mask_generation_autoencoder_path: models/mask_generation_autoencoder.pt\n", - "trained_mask_generation_diffusion_path: models/mask_generation_diffusion_unet.pt\n", - "all_mask_files_base_dir: /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000\n", - "all_mask_files_json: ./configs/candidate_masks_flexible_size_and_spacing_3000.json\n", - "all_anatomy_size_conditions_json: ./configs/all_anatomy_size_condtions.json\n", - "label_dict_json: ./configs/label_dict.json\n", - "label_dict_remap_json: ./configs/label_dict_124_to_132.json\n", - "Global config variables have been loaded.\n" + "[2025-03-19 05:06:22.493][ INFO](notebook) - output_dir: output\n", + "[2025-03-19 05:06:22.494][ INFO](notebook) - trained_autoencoder_path: models/autoencoder_epoch273.pt\n", + "[2025-03-19 05:06:22.494][ INFO](notebook) - trained_diffusion_path: models/diff_unet_3d_rflow.pt\n", + "[2025-03-19 05:06:22.495][ INFO](notebook) - trained_controlnet_path: models/controlnet_3d_rflow.pt\n", + "[2025-03-19 05:06:22.496][ INFO](notebook) - trained_mask_generation_autoencoder_path: models/mask_generation_autoencoder.pt\n", + "[2025-03-19 05:06:22.497][ INFO](notebook) - trained_mask_generation_diffusion_path: models/mask_generation_diffusion_unet.pt\n", + "[2025-03-19 05:06:22.497][ INFO](notebook) - all_mask_files_base_dir: temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000\n", + "[2025-03-19 05:06:22.498][ INFO](notebook) - all_mask_files_json: ./configs/candidate_masks_flexible_size_and_spacing_4000.json\n", + "[2025-03-19 05:06:22.498][ INFO](notebook) - all_anatomy_size_conditions_json: ./configs/all_anatomy_size_condtions.json\n", + "[2025-03-19 05:06:22.499][ INFO](notebook) - label_dict_json: ./configs/label_dict.json\n", + "[2025-03-19 05:06:22.500][ INFO](notebook) - label_dict_remap_json: ./configs/label_dict_124_to_132.json\n", + "[2025-03-19 05:06:22.501][ INFO](notebook) - Global config variables have been loaded.\n" ] } ], "source": [ "args = argparse.Namespace()\n", "\n", - "environment_file = \"./configs/environment.json\"\n", + "if maisi_version == \"maisi3d-ddpm\":\n", + " environment_file = \"./configs/environment_maisi3d-ddpm.json\"\n", + "elif maisi_version == \"maisi3d-rflow\":\n", + " environment_file = \"./configs/environment_maisi3d-rflow.json\"\n", + "else:\n", + " raise ValueError(f\"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}.\")\n", + "\n", "with open(environment_file, \"r\") as f:\n", " env_dict = json.load(f)\n", "for k, v in env_dict.items():\n", " # Update the path to the downloaded dataset in MONAI_DATA_DIRECTORY\n", " val = v if \"datasets/\" not in v else os.path.join(root_dir, v)\n", " setattr(args, k, val)\n", - " print(f\"{k}: {val}\")\n", - "print(\"Global config variables have been loaded.\")" + " logger.info(f\"{k}: {val}\")\n", + "logger.info(\"Global config variables have been loaded.\")" ] }, { @@ -269,7 +343,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 6, "id": "533414f3-bef5-49f7-b082-f803b5e494bf", "metadata": {}, "outputs": [ @@ -277,36 +351,35 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:root:`controllable_anatomy_size` is empty.\n", - "We will synthesize based on `body_region`: (['abdomen']) and `anatomy_list`: (['liver', 'hepatic tumor']).\n", - "INFO:root:The generate results will have voxel size to be [1.5, 1.5, 2.0]mm, volume size to be [256, 256, 256].\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "num_output_samples: 1\n", - "body_region: ['abdomen']\n", - "anatomy_list: ['liver', 'hepatic tumor']\n", - "controllable_anatomy_size: []\n", - "num_inference_steps: 1000\n", - "mask_generation_num_inference_steps: 1000\n", - "output_size: [256, 256, 256]\n", - "image_output_ext: .nii.gz\n", - "label_output_ext: .nii.gz\n", - "spacing: [1.5, 1.5, 2.0]\n", - "autoencoder_sliding_window_infer_size: [48, 48, 48]\n", - "autoencoder_sliding_window_infer_overlap: 0.25\n", - "Network definition and inference inputs have been loaded.\n" + "[2025-03-19 05:06:22.508][ INFO](notebook) - num_output_samples: 1\n", + "[2025-03-19 05:06:22.509][ INFO](notebook) - body_region: ['chest']\n", + "[2025-03-19 05:06:22.509][ INFO](notebook) - anatomy_list: ['lung tumor']\n", + "[2025-03-19 05:06:22.510][ INFO](notebook) - controllable_anatomy_size: []\n", + "[2025-03-19 05:06:22.510][ INFO](notebook) - num_inference_steps: 30\n", + "[2025-03-19 05:06:22.512][ INFO](notebook) - mask_generation_num_inference_steps: 1000\n", + "[2025-03-19 05:06:22.512][ INFO](notebook) - output_size: [256, 256, 256]\n", + "[2025-03-19 05:06:22.513][ INFO](notebook) - image_output_ext: .nii.gz\n", + "[2025-03-19 05:06:22.514][ INFO](notebook) - label_output_ext: .nii.gz\n", + "[2025-03-19 05:06:22.514][ INFO](notebook) - spacing: [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:22.515][ INFO](notebook) - autoencoder_sliding_window_infer_size: [48, 48, 48]\n", + "[2025-03-19 05:06:22.515][ INFO](notebook) - autoencoder_sliding_window_infer_overlap: 0.6666\n", + "[2025-03-19 05:06:22.516][ INFO](notebook) - controlnet: $@controlnet_def\n", + "[2025-03-19 05:06:22.517][ INFO](notebook) - diffusion_unet: $@diffusion_unet_def\n", + "[2025-03-19 05:06:22.518][ INFO](notebook) - autoencoder: $@autoencoder_def\n", + "[2025-03-19 05:06:22.518][ INFO](notebook) - mask_generation_autoencoder: $@mask_generation_autoencoder_def\n", + "[2025-03-19 05:06:22.519][ INFO](notebook) - mask_generation_diffusion: $@mask_generation_diffusion_def\n", + "[2025-03-19 05:06:22.520][ INFO](notebook) - modality: 1\n", + "[2025-03-19 05:06:22.521][ INFO](root) - `controllable_anatomy_size` is empty.\n", + "We will synthesize based on `body_region`: (['chest']) and `anatomy_list`: (['lung tumor']).\n", + "[2025-03-19 05:06:22.522][ INFO](root) - The generate results will have voxel size to be [1.5, 1.5, 2.0]mm, volume size to be [256, 256, 256].\n", + "[2025-03-19 05:06:22.522][ INFO](notebook) - Network definition and inference inputs have been loaded.\n" ] } ], "source": [ - "config_file = \"./configs/config_maisi.json\"\n", - "with open(config_file, \"r\") as f:\n", - " config_dict = json.load(f)\n", - "for k, v in config_dict.items():\n", + "with open(model_def_path, \"r\") as f:\n", + " model_def = json.load(f)\n", + "for k, v in model_def.items():\n", " setattr(args, k, v)\n", "\n", "# check the format of inference inputs\n", @@ -315,7 +388,7 @@ " config_infer_dict = json.load(f)\n", "for k, v in config_infer_dict.items():\n", " setattr(args, k, v)\n", - " print(f\"{k}: {v}\")\n", + " logger.info(f\"{k}: {v}\")\n", "\n", "check_input(\n", " args.body_region,\n", @@ -326,7 +399,7 @@ " args.controllable_anatomy_size,\n", ")\n", "latent_shape = [args.latent_channels, args.output_size[0] // 4, args.output_size[1] // 4, args.output_size[2] // 4]\n", - "print(\"Network definition and inference inputs have been loaded.\")" + "logger.info(\"Network definition and inference inputs have been loaded.\")" ] }, { @@ -339,7 +412,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 7, "id": "87ba613d-a2f5-4afc-95df-65ad21fafedd", "metadata": {}, "outputs": [], @@ -360,7 +433,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "id": "d499f7b1", "metadata": { "lines_to_next_cell": 2 @@ -370,8 +443,14 @@ "name": "stdout", "output_type": "stream", "text": [ - "2024-09-30 06:34:48,434 - INFO - 'dst' model updated: 158 of 206 variables.\n", - "All the trained model weights have been loaded.\n" + "2025-03-19 05:06:28,553 - INFO - 'dst' model updated: 180 of 231 variables.\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "[2025-03-19 05:06:30.944][ INFO](notebook) - All the trained model weights have been loaded.\n" ] } ], @@ -404,7 +483,7 @@ "mask_generation_diffusion_unet.load_state_dict(checkpoint_mask_generation_diffusion_unet[\"unet_state_dict\"])\n", "mask_generation_scale_factor = checkpoint_mask_generation_diffusion_unet[\"scale_factor\"]\n", "\n", - "print(\"All the trained model weights have been loaded.\")" + "logger.info(\"All the trained model weights have been loaded.\")" ] }, { @@ -417,7 +496,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "id": "8685da6e", "metadata": {}, "outputs": [ @@ -425,7 +504,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:root:LDM sampler initialized.\n" + "[2025-03-19 05:06:30.969][ INFO](root) - LDM sampler initialized.\n" ] } ], @@ -456,6 +535,7 @@ " image_output_ext=args.image_output_ext,\n", " label_output_ext=args.label_output_ext,\n", " spacing=args.spacing,\n", + " modality=args.modality,\n", " num_inference_steps=args.num_inference_steps,\n", " mask_generation_num_inference_steps=args.mask_generation_num_inference_steps,\n", " random_seed=args.random_seed,\n", @@ -477,95 +557,96 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 10, "id": "271f91bf-1c55-46e2-ae56-8677cd8eb81f", "metadata": { "scrolled": true }, "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "The generated image/mask pairs will be saved in output.\n", - "Extracting /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000.zip to /workspace/data/datasets\n", - "2024-09-30 06:34:50,652 - INFO - Writing into directory: /workspace/data/datasets.\n" - ] - }, { "name": "stderr", "output_type": "stream", "text": [ - "INFO:root:Resample mask file to get desired output size and spacing\n" + "[2025-03-19 05:06:30.974][ INFO](notebook) - The generated image/mask pairs will be saved in output.\n", + "[2025-03-19 05:06:31.018][ INFO](root) - Resample mask file to get desired output size and spacing\n", + "[2025-03-19 05:06:32.950][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:32.953][ INFO](root) - Resize Spacing: [tensor(0.7988, dtype=torch.float64), tensor(0.7988, dtype=torch.float64), tensor(1.1016, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:32.954][ INFO](root) - Output size: [512, 512, 384] -> [256, 256, 256]\n", + "[2025-03-19 05:06:35.876][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:35.878][ INFO](root) - Resize Spacing: [tensor(0.7031, dtype=torch.float64), tensor(0.7031, dtype=torch.float64), tensor(1.4795, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:35.879][ INFO](root) - Output size: [512, 512, 256] -> [256, 256, 256]\n", + "[2025-03-19 05:06:38.564][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:38.566][ INFO](root) - Resize Spacing: [tensor(0.7617, dtype=torch.float64), tensor(0.7617, dtype=torch.float64), tensor(1.2939, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:38.567][ INFO](root) - Output size: [512, 512, 256] -> [256, 256, 256]\n", + "[2025-03-19 05:06:41.321][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:41.324][ INFO](root) - Resize Spacing: [tensor(0.7031, dtype=torch.float64), tensor(0.7031, dtype=torch.float64), tensor(1.4209, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:41.325][ INFO](root) - Output size: [512, 512, 256] -> [256, 256, 256]\n", + "[2025-03-19 05:06:46.041][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:46.043][ INFO](root) - Resize Spacing: [tensor(0.7422, dtype=torch.float64), tensor(0.7422, dtype=torch.float64), tensor(0.5752, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:46.044][ INFO](root) - Output size: [512, 512, 640] -> [256, 256, 256]\n", + "[2025-03-19 05:06:47.854][ INFO](root) - Images will be generated based on [{'mask_file': {'pseudo_label': 'temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000/./Task06/labelsTr/lung_070_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [1, 0, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}, 'if_aug': True}, {'mask_file': {'pseudo_label': 'temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000/./Task06/labelsTr/lung_031_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [0, 1, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}, 'if_aug': True}, {'mask_file': {'pseudo_label': 'temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000/./Task06/labelsTr/lung_075_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [1, 0, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}, 'if_aug': True}, {'mask_file': {'pseudo_label': 'temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000/./Task06/labelsTr/lung_014_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [0, 1, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}, 'if_aug': True}, {'mask_file': {'pseudo_label': 'temp_work_dir_inference_demo/datasets/all_masks_flexible_size_and_spacing_4000/./Task06/labelsTr/lung_049_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [1, 0, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}, 'if_aug': True}].\n", + "[2025-03-19 05:06:47.855][ INFO](root) - ---- Start preparing masks... ----\n", + "[2025-03-19 05:06:49.096][ INFO](root) - Resampling mask to target shape and spacing\n", + "[2025-03-19 05:06:49.100][ INFO](root) - Resize Spacing: [tensor(0.7617, dtype=torch.float64), tensor(0.7617, dtype=torch.float64), tensor(1.2939, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", + "[2025-03-19 05:06:49.100][ INFO](root) - Output size: [512, 512, 256] -> [256, 256, 256]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ - "Unzipped /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000.zip to /workspace/data/datasets/all_masks_flexible_size_and_spacing_3000.\n" + "augmenting lung tumor\n", + "28\n", + "metatensor(180., device='cuda:0') | metatensor(547.4000, device='cuda:0')\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.7988, dtype=torch.float64), tensor(0.7988, dtype=torch.float64), tensor(1.9062, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n", - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.7852, dtype=torch.float64), tensor(0.7852, dtype=torch.float64), tensor(1.9336, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n", - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.8027, dtype=torch.float64), tensor(0.8027, dtype=torch.float64), tensor(1.8672, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n", - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.9062, dtype=torch.float64), tensor(0.9062, dtype=torch.float64), tensor(2.3438, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n", - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.9551, dtype=torch.float64), tensor(0.9551, dtype=torch.float64), tensor(2.4805, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n", - "INFO:root:Images will be generated based on [{'mask_file': {'pseudo_label': '/workspace/data/datasets/all_masks_flexible_size_and_spacing_3000/./Task03/labelsTr/liver_56_133combined_aug_wbdm.nii.gz', 'spacing': [1.5, 1.5, 2.0], 'dim': [256, 256, 256], 'top_region_index': [0, 1, 0, 0], 'bottom_region_index': [0, 0, 0, 1]}, 'if_aug': True}].\n", - "INFO:root:---- Start preparing masks... ----\n", - "INFO:root:Resampling mask to target shape and spacing\n", - "INFO:root:Resize Spacing: [tensor(0.8027, dtype=torch.float64), tensor(0.8027, dtype=torch.float64), tensor(1.8672, dtype=torch.float64)] -> [1.5, 1.5, 2.0]\n", - "INFO:root:Output size: [512, 512, 256] -> [256, 256, 256]\n" + "[2025-03-19 05:06:52.345][ INFO](root) - ---- Mask preparation time: 4.489694595336914 seconds ----\n", + "[2025-03-19 05:06:52.363][ INFO](root) - ---- Start generating latent features... ----\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ - "augmenting liver tumor\n" + "metatensor(687., device='cuda:0') | metatensor(547.4000, device='cuda:0')\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ - "INFO:root:---- Mask preparation time: 10.283801794052124 seconds ----\n", - "INFO:root:---- Start generating latent features... ----\n", - "100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1000/1000 [02:15<00:00, 7.41it/s]\n", - "INFO:root:---- Latent features generation time: 135.05768537521362 seconds ----\n", - "INFO:root:---- Start decoding latent features into images... ----\n", - "100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 8/8 [00:08<00:00, 1.09s/it]\n", - "INFO:root:---- Image decoding time: 8.780551671981812 seconds ----\n" + "100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 30/30 [00:06<00:00, 4.58it/s]\n", + "[2025-03-19 05:06:58.962][ INFO](root) - ---- DM/ControlNet Latent features generation time: 6.598896265029907 seconds ----\n", + "[2025-03-19 05:06:59.051][ INFO](root) - ---- Start decoding latent features into images... ----\n", + "100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 8/8 [00:11<00:00, 1.42s/it]\n", + "[2025-03-19 05:07:10.494][ INFO](root) - ---- Image VAE decoding time: 11.442715883255005 seconds ----\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ - "2024-09-30 06:38:43,064 INFO image_writer.py:197 - writing: output/sample_20240930_063843_052831_image.nii.gz\n", - "2024-09-30 06:38:44,661 INFO image_writer.py:197 - writing: output/sample_20240930_063843_052831_label.nii.gz\n", - "MAISI image/mask generation finished\n" + "1 5\n", + "2025-03-19 05:07:10,994 INFO image_writer.py:197 - writing: output/sample_20250319_050710_975233_image.nii.gz\n", + "2025-03-19 05:07:12,461 INFO image_writer.py:197 - writing: output/sample_20250319_050710_975233_label.nii.gz\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "[2025-03-19 05:07:13.422][ INFO](notebook) - MAISI image/mask generation finished\n" ] } ], "source": [ - "print(f\"The generated image/mask pairs will be saved in {args.output_dir}.\")\n", + "logger.info(f\"The generated image/mask pairs will be saved in {args.output_dir}.\")\n", "output_filenames = ldm_sampler.sample_multiple_images(args.num_output_samples)\n", - "print(\"MAISI image/mask generation finished\")" + "logger.info(\"MAISI image/mask generation finished\")" ] }, { @@ -578,22 +659,20 @@ }, { "cell_type": "code", - "execution_count": 10, - "id": "e0453d9f-1614-4c84-aef1-77b6339d8c12", - "metadata": { - "scrolled": true - }, + "execution_count": 11, + "id": "dfd2ebf9-04f9-498f-982e-9daf44602bee", + "metadata": {}, "outputs": [ { - "name": "stdout", + "name": "stderr", "output_type": "stream", "text": [ - "Visualizing output/sample_20240930_063843_052831_image.nii.gz and output/sample_20240930_063843_052831_label.nii.gz...\n" + "[2025-03-19 05:07:13.430][ INFO](notebook) - Visualizing output/sample_20250319_050710_975233_image.nii.gz and output/sample_20250319_050710_975233_label.nii.gz...\n" ] }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -603,7 +682,7 @@ }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -615,7 +694,7 @@ "source": [ "visualize_image_filename = output_filenames[0][0]\n", "visualize_mask_filename = output_filenames[0][1]\n", - "print(f\"Visualizing {visualize_image_filename} and {visualize_mask_filename}...\")\n", + "logger.info(f\"Visualizing {visualize_image_filename} and {visualize_mask_filename}...\")\n", "\n", "# load image/mask pairs\n", "loader = LoadImage(image_only=True, ensure_channel_first=True)\n", @@ -624,7 +703,7 @@ "mask_volume = orientation(loader(visualize_mask_filename)).to(torch.uint8)\n", "\n", "# visualize for CT HU intensity between [-200, 500]\n", - "image_volume = torch.clip(image_volume, -200, 500)\n", + "image_volume = torch.clip(image_volume, -1000, 300)\n", "image_volume = image_volume - torch.min(image_volume)\n", "image_volume = image_volume / torch.max(image_volume)\n", "\n", @@ -651,7 +730,7 @@ "formats": "py:percent,ipynb" }, "kernelspec": { - "display_name": "Python 3", + "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, diff --git a/generation/maisi/maisi_train_controlnet_tutorial.ipynb b/generation/maisi/maisi_train_controlnet_tutorial.ipynb index 37becc2e43..565adb7cd4 100644 --- a/generation/maisi/maisi_train_controlnet_tutorial.ipynb +++ b/generation/maisi/maisi_train_controlnet_tutorial.ipynb @@ -26,7 +26,9 @@ "\n", "![Generated image examples and input mask](https://developer.download.nvidia.com/assets/Clara/Images/monai_maisi_ct_generative_example_synthetic_data.png)\n", "\n", - "In this notebook, we detail the procedure for training a 3D ControlNet to generate high-dimensional 3D medical images. Due to the potential for out-of-memory issues on most GPUs when generating large images (e.g., those with dimensions of 512 x 512 x 512 or greater), we have structured the training process into two primary steps: 1) preparing training data, 2) training config preparation, and 3) launch training of 3D ControlNet. The subsequent sections will demonstrate the entire process using a simulated dataset. We also provide the real preprocessed dataset used in the finetuning config `environment_maisi_controlnet_train.json`. More instructions about how to preprocess real data can be found in the [README](./data/README.md) in `data` folder.\n" + "In this notebook, we detail the procedure for training a 3D ControlNet to generate high-dimensional 3D medical images. Due to the potential for out-of-memory issues on most GPUs when generating large images (e.g., those with dimensions of 512 x 512 x 512 or greater), we have structured the training process into two primary steps: 1) preparing training data, 2) training config preparation, and 3) launch training of 3D ControlNet. The subsequent sections will demonstrate the entire process using a simulated dataset. We also provide the real preprocessed dataset used in the finetuning config `environment_maisi_controlnet_train.json`. More instructions about how to preprocess real data can be found in the [README](./data/README.md) in `data` folder.\n", + "\n", + "`[Release Note (March 2025)]:` We are excited to announce the new MAISI Version `'maisi3d-rflow'`. Compared with the previous version `'maisi3d-ddpm'`, it accelerated latent diffusion model inference by 33x. Please see the detailed difference in the following section." ] }, { @@ -57,46 +59,38 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 2, "id": "e3bf0346", "metadata": {}, "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/usr/local/lib/python3.10/dist-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", - " from .autonotebook import tqdm as notebook_tqdm\n" - ] - }, { "name": "stdout", "output_type": "stream", "text": [ - "MONAI version: 1.4.0rc12\n", - "Numpy version: 1.24.4\n", - "Pytorch version: 2.5.0a0+872d972e41.nv24.08\n", + "MONAI version: 1.4.1rc1+32.g34f37973\n", + "Numpy version: 1.26.4\n", + "Pytorch version: 2.5.0+cu124\n", "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n", - "MONAI rev id: 76ef9f40c8da626928238c91eacddc789b0b4530\n", - "MONAI __file__: /workspace/Code/MONAI/monai/__init__.py\n", + "MONAI rev id: 34f379735c5e18e7f809453eb1b3606c225c788b\n", + "MONAI __file__: /localhome//.local/lib/python3.10/site-packages/monai/__init__.py\n", "\n", "Optional dependencies:\n", "Pytorch Ignite version: 0.4.11\n", "ITK version: 5.4.0\n", - "Nibabel version: 5.2.1\n", - "scikit-image version: 0.23.2\n", - "scipy version: 1.14.0\n", - "Pillow version: 10.4.0\n", - "Tensorboard version: 2.16.2\n", + "Nibabel version: 5.3.2\n", + "scikit-image version: 0.24.0\n", + "scipy version: 1.14.1\n", + "Pillow version: 11.0.0\n", + "Tensorboard version: 2.18.0\n", "gdown version: 5.2.0\n", - "TorchVision version: 0.20.0a0\n", + "TorchVision version: 0.20.0+cu124\n", "tqdm version: 4.66.5\n", "lmdb version: 1.5.1\n", - "psutil version: 6.0.0\n", - "pandas version: 2.2.2\n", + "psutil version: 6.1.0\n", + "pandas version: 2.2.3\n", "einops version: 0.8.0\n", "transformers version: 4.40.2\n", - "mlflow version: 2.16.2\n", + "mlflow version: 2.17.1\n", "pynrrd version: 1.0.0\n", "clearml version: 1.16.5rc2\n", "\n", @@ -124,6 +118,47 @@ "logger = setup_logging(\"notebook\")" ] }, + { + "cell_type": "markdown", + "id": "b16b92a4-2039-42b7-bf77-68851d25701b", + "metadata": {}, + "source": [ + "## Set up the MAISI version\n", + "\n", + "Choose between `'maisi3d-ddpm'` and `'maisi3d-rflow'`. The differences are:\n", + "- The maisi version `'maisi3d-ddpm'` uses basic noise scheduler DDPM. `'maisi3d-rflow'` uses Rectified Flow scheduler, can be 33 times faster during inference.\n", + "- The maisi version `'maisi3d-ddpm'` requires training images to be labeled with body regions (`\"top_region_index\"` and `\"bottom_region_index\"`), while `'maisi3d-rflow'` does not have such requirement. In other words, it is easier to prepare training data for `'maisi3d-rflow'`.\n", + "- For the released model weights, `'maisi3d-rflow'` can generate images with better quality for head region and small output volumes, and comparable quality for other cases compared with `'maisi3d-ddpm'`." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "3d233abe-d69c-4b57-9655-33c2c3da6c96", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "[2025-03-14 16:29:13.938][ INFO](notebook) - MAISI version is maisi3d-rflow, whether to use body_region is False\n" + ] + } + ], + "source": [ + "maisi_version = \"maisi3d-rflow\"\n", + "if maisi_version == \"maisi3d-ddpm\":\n", + " model_def_path = \"./configs/config_maisi3d-ddpm.json\"\n", + "elif maisi_version == \"maisi3d-rflow\":\n", + " model_def_path = \"./configs/config_maisi3d-rflow.json\"\n", + "else:\n", + " raise ValueError(f\"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}.\")\n", + "with open(model_def_path, \"r\") as f:\n", + " model_def = json.load(f)\n", + "include_body_region = model_def[\"include_body_region\"]\n", + "logger.info(f\"MAISI version is {maisi_version}, whether to use body_region is {include_body_region}\")" + ] + }, { "cell_type": "markdown", "id": "671e7f10", @@ -154,7 +189,7 @@ }, { "cell_type": "code", - "execution_count": 2, + "execution_count": 4, "id": "fc32a7fe", "metadata": {}, "outputs": [], @@ -169,8 +204,6 @@ " \"fold\": 0, # fold index for cross validation, fold 0 is used for training\n", " \"dim\": sim_dim, # the dimension of image\n", " \"spacing\": [1.5, 1.5, 1.5], # the spacing of image\n", - " \"top_region_index\": [0, 1, 0, 0], # the top region index of the image\n", - " \"bottom_region_index\": [0, 0, 0, 1], # the bottom region index of the image\n", " },\n", " {\n", " \"image\": \"tr_image_002_emb.nii.gz\",\n", @@ -178,8 +211,6 @@ " \"fold\": 1,\n", " \"dim\": sim_dim,\n", " \"spacing\": [1.5, 1.5, 1.5],\n", - " \"top_region_index\": [0, 1, 0, 0],\n", - " \"bottom_region_index\": [0, 0, 0, 1],\n", " },\n", " {\n", " \"image\": \"tr_image_003_emb.nii.gz\",\n", @@ -187,11 +218,14 @@ " \"fold\": 1,\n", " \"dim\": sim_dim,\n", " \"spacing\": [1.5, 1.5, 1.5],\n", - " \"top_region_index\": [0, 1, 0, 0],\n", - " \"bottom_region_index\": [0, 0, 0, 1],\n", " },\n", " ]\n", - "}" + "}\n", + "if include_body_region:\n", + " for i in range(len(sim_datalist[\"training\"])):\n", + " # body region index\n", + " sim_datalist[\"training\"][i][\"top_region_index\"] = [0, 1, 0, 0] # the top region index of the image\n", + " sim_datalist[\"training\"][i][\"bottom_region_index\"] = [0, 0, 0, 1] # the bottom region index of the image" ] }, { @@ -206,7 +240,7 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 5, "id": "1b199078", "metadata": {}, "outputs": [ @@ -214,9 +248,10 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Generated simulated images.\n", - "INFO:notebook:img_emb shape: (64, 64, 32, 4)\n", - "INFO:notebook:label shape: (256, 256, 128)\n" + "[2025-03-14 16:29:13.952][ INFO](notebook) - Save data list json file to ./temp_work_dir_controlnet_train_demo/sim_datalist.json\n", + "[2025-03-14 16:29:16.033][ INFO](notebook) - Generated simulated images.\n", + "[2025-03-14 16:29:16.034][ INFO](notebook) - img_emb shape: (64, 64, 32, 4)\n", + "[2025-03-14 16:29:16.035][ INFO](notebook) - label shape: (256, 256, 128)\n" ] } ], @@ -232,6 +267,7 @@ "datalist_file = os.path.join(work_dir, \"sim_datalist.json\")\n", "with open(datalist_file, \"w\") as f:\n", " json.dump(sim_datalist, f, indent=4)\n", + "logger.info(f\"Save data list json file to {datalist_file}\")\n", "\n", "for d in sim_datalist[\"training\"]:\n", " # The image embedding is downsampled twice by Autoencoder.\n", @@ -280,7 +316,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 6, "id": "6c7b434c", "metadata": {}, "outputs": [ @@ -288,15 +324,14 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:files and folders under work_dir: ['config_maisi.json', 'sim_dataroot', 'sim_datalist.json', 'models', 'outputs', 'environment_maisi_controlnet_train.json', 'config_maisi_controlnet_train.json'].\n", - "INFO:notebook:number of GPUs: 1.\n" + "[2025-03-14 16:29:16.049][ INFO](notebook) - files and folders under work_dir: ['config_maisi.json', 'models', 'config_maisi_controlnet_train.json', 'outputs', 'sim_dataroot', '.ipynb_checkpoints', 'environment_maisi_controlnet_train.json', 'sim_datalist.json'].\n", + "[2025-03-14 16:29:16.050][ INFO](notebook) - number of GPUs: 1.\n" ] } ], "source": [ "env_config_path = \"./configs/environment_maisi_controlnet_train.json\"\n", "train_config_path = \"./configs/config_maisi_controlnet_train.json\"\n", - "model_def_path = \"./configs/config_maisi.json\"\n", "\n", "# Load environment configuration, model configuration and model definition\n", "with open(env_config_path, \"r\") as f:\n", @@ -367,7 +402,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 7, "id": "95ea6972", "metadata": {}, "outputs": [], @@ -427,7 +462,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 8, "id": "ade6389d", "metadata": {}, "outputs": [ @@ -435,30 +470,30 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Training the model...\n" + "[2025-03-14 16:29:16.061][ INFO](notebook) - Training the model...\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ - "2024-09-24 02:33:40,881 - INFO - 'dst' model updated: 158 of 206 variables.\n", - "\n", - "INFO:maisi.controlnet.training:Number of GPUs: 2\n", - "INFO:maisi.controlnet.training:World_size: 1\n", - "INFO:maisi.controlnet.training:trained diffusion model is not loaded.\n", - "INFO:maisi.controlnet.training:set scale_factor -> 1.0.\n", - "INFO:maisi.controlnet.training:train controlnet model from scratch.\n", - "INFO:maisi.controlnet.training:total number of training steps: 4.0.\n", - "INFO:maisi.controlnet.training:\n", - "[Epoch 1/2] [Batch 1/2] [LR: 0.00000563] [loss: 0.7981] ETA: 0:00:01.501654 \n", - "INFO:maisi.controlnet.training:\n", - "[Epoch 1/2] [Batch 2/2] [LR: 0.00000250] [loss: 0.7976] ETA: 0:00:00 \n", - "INFO:maisi.controlnet.training:best loss -> 0.7978459596633911.\n", - "INFO:maisi.controlnet.training:\n", - "[Epoch 2/2] [Batch 1/2] [LR: 0.00000063] [loss: 0.7982] ETA: 0:00:01.988772 \n", - "INFO:maisi.controlnet.training:\n", - "[Epoch 2/2] [Batch 2/2] [LR: 0.00000000] [loss: 0.7998] ETA: 0:00:00 \n", + "[2025-03-14 16:29:23.336][ INFO](maisi.controlnet.training) - Number of GPUs: 8\n", + "[2025-03-14 16:29:23.336][ INFO](maisi.controlnet.training) - World_size: 1\n", + "[2025-03-14 16:29:24.771][ INFO](maisi.controlnet.training) - trained diffusion model is not loaded.\n", + "[2025-03-14 16:29:24.771][ INFO](maisi.controlnet.training) - set scale_factor -> 1.0.\n", + "2025-03-14 16:29:25,271 - INFO - 'dst' model updated: 180 of 231 variables.\n", + "[2025-03-14 16:29:25.277][ INFO](maisi.controlnet.training) - train controlnet model from scratch.\n", + "[2025-03-14 16:29:25.300][ INFO](maisi.controlnet.training) - total number of training steps: 4.0.\n", + "[2025-03-14 16:29:26.826][ INFO](maisi.controlnet.training) -\n", + "[Epoch 1/2] [Batch 1/2] [LR: 0.00000563] [loss: 0.8278] ETA: 0:00:01.523338\n", + "[2025-03-14 16:29:26.974][ INFO](maisi.controlnet.training) -\n", + "[Epoch 1/2] [Batch 2/2] [LR: 0.00000250] [loss: 0.8289] ETA: 0:00:00\n", + "[2025-03-14 16:29:27.585][ INFO](maisi.controlnet.training) - best loss -> 0.8283329606056213.\n", + "[2025-03-14 16:29:28.909][ INFO](maisi.controlnet.training) -\n", + "[Epoch 2/2] [Batch 1/2] [LR: 0.00000063] [loss: 0.8288] ETA: 0:00:01.934548\n", + "[2025-03-14 16:29:29.052][ INFO](maisi.controlnet.training) -\n", + "[Epoch 2/2] [Batch 2/2] [LR: 0.00000000] [loss: 0.8277] ETA: 0:00:00\n", + "[2025-03-14 16:29:29.716][ INFO](maisi.controlnet.training) - best loss -> 0.8282470703125.\n", "\n" ] } @@ -493,7 +528,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 9, "id": "936360c8", "metadata": {}, "outputs": [ @@ -501,32 +536,32 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Inference...\n" + "[2025-03-14 16:29:32.229][ INFO](notebook) - Inference...\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ - "2024-09-24 02:34:03,472 - INFO - 'dst' model updated: 158 of 206 variables.\n", - "2024-09-24 02:34:06,052 INFO image_writer.py:197 - writing: temp_work_dir_controlnet_train_demo/outputs/sample_20240924_023406_038072_image.nii.gz\n", - "2024-09-24 02:34:06,437 INFO image_writer.py:197 - writing: temp_work_dir_controlnet_train_demo/outputs/sample_20240924_023406_038072_label.nii.gz\n", + "[2025-03-14 16:29:39.519][ INFO](maisi.controlnet.infer) - Number of GPUs: 8\n", + "[2025-03-14 16:29:39.519][ INFO](maisi.controlnet.infer) - World_size: 1\n", + "[2025-03-14 16:29:39.990][ INFO](maisi.controlnet.infer) - trained autoencoder model is not loaded.\n", + "[2025-03-14 16:29:41.213][ INFO](maisi.controlnet.infer) - trained diffusion model is not loaded.\n", + "[2025-03-14 16:29:41.213][ INFO](maisi.controlnet.infer) - set scale_factor -> 1.0.\n", + "2025-03-14 16:29:41,716 - INFO - 'dst' model updated: 180 of 231 variables.\n", + "[2025-03-14 16:29:41.721][ INFO](maisi.controlnet.infer) - trained controlnet is not loaded.\n", + "[2025-03-14 16:29:42.102][ INFO](root) - `controllable_anatomy_size` is not provided.\n", + "[2025-03-14 16:29:42.104][ INFO](root) - ---- Start generating latent features... ----\n", + "[2025-03-14 16:29:42.670][ INFO](root) - ---- DM/ControlNet Latent features generation time: 0.565190315246582 seconds ----\n", + "[2025-03-14 16:29:42.672][ INFO](root) - ---- Start decoding latent features into images... ----\n", + "[2025-03-14 16:29:43.314][ INFO](root) - ---- Image VAE decoding time: 0.6416211128234863 seconds ----\n", + "2025-03-14 16:29:43,602 INFO image_writer.py:197 - writing: temp_work_dir_controlnet_train_demo/outputs/sample_20250314_162943_586788_image.nii.gz\n", + "2025-03-14 16:29:43,940 INFO image_writer.py:197 - writing: temp_work_dir_controlnet_train_demo/outputs/sample_20250314_162943_586788_label.nii.gz\n", "\n", - "INFO:maisi.controlnet.infer:Number of GPUs: 2\n", - "INFO:maisi.controlnet.infer:World_size: 1\n", - "INFO:maisi.controlnet.infer:trained autoencoder model is not loaded.\n", - "INFO:maisi.controlnet.infer:trained diffusion model is not loaded.\n", - "INFO:maisi.controlnet.infer:set scale_factor -> 1.0.\n", - "INFO:maisi.controlnet.infer:trained controlnet is not loaded.\n", - "INFO:root:`controllable_anatomy_size` is not provided.\n", - "INFO:root:---- Start generating latent features... ----\n", "\n", - " 0%| | 0/1 [00:00/.local/lib/python3.10/site-packages/monai/__init__.py\n", + "\n", + "Optional dependencies:\n", + "Pytorch Ignite version: 0.4.11\n", + "ITK version: 5.4.0\n", + "Nibabel version: 5.3.2\n", + "scikit-image version: 0.24.0\n", + "scipy version: 1.14.1\n", + "Pillow version: 11.0.0\n", + "Tensorboard version: 2.18.0\n", + "gdown version: 5.2.0\n", + "TorchVision version: 0.20.0+cu124\n", + "tqdm version: 4.66.5\n", + "lmdb version: 1.5.1\n", + "psutil version: 6.1.0\n", + "pandas version: 2.2.3\n", + "einops version: 0.8.0\n", + "transformers version: 4.40.2\n", + "mlflow version: 2.17.1\n", + "pynrrd version: 1.0.0\n", + "clearml version: 1.16.5rc2\n", + "\n", + "For details about installing the optional dependencies, please visit:\n", + " https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n", + "\n" + ] + } + ], "source": [ - "from scripts.diff_model_setting import setup_logging\n", "import copy\n", "import os\n", "import json\n", "import numpy as np\n", "import nibabel as nib\n", "import subprocess\n", + "from IPython.display import Image, display\n", "\n", "from monai.apps import download_url\n", "from monai.data import create_test_image_3d\n", "from monai.config import print_config\n", "\n", - "from IPython.display import Image, display\n", + "from scripts.diff_model_setting import setup_logging\n", "\n", "print_config()\n", "\n", "logger = setup_logging(\"notebook\")" ] }, + { + "cell_type": "markdown", + "id": "c2389853", + "metadata": {}, + "source": [ + "## Set up the MAISI version\n", + "\n", + "Choose between `'maisi3d-ddpm'` and `'maisi3d-rflow'`. The differences are:\n", + "- The maisi version `'maisi3d-ddpm'` uses basic noise scheduler DDPM. `'maisi3d-rflow'` uses Rectified Flow scheduler, can be 33 times faster during inference.\n", + "- The maisi version `'maisi3d-ddpm'` requires training images to be labeled with body regions (`\"top_region_index\"` and `\"bottom_region_index\"`), while `'maisi3d-rflow'` does not have such requirement. In other words, it is easier to prepare training data for `'maisi3d-rflow'`.\n", + "- For the released model weights, `'maisi3d-rflow'` can generate images with better quality for head region and small output volumes, and comparable quality for other cases compared with `'maisi3d-ddpm'`." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "31684f74", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "[2025-03-14 16:14:21.679][ INFO](notebook) - MAISI version is maisi3d-rflow, whether to use body_region is False\n" + ] + } + ], + "source": [ + "maisi_version = \"maisi3d-rflow\"\n", + "if maisi_version == \"maisi3d-ddpm\":\n", + " model_def_path = \"./configs/config_maisi3d-ddpm.json\"\n", + "elif maisi_version == \"maisi3d-rflow\":\n", + " model_def_path = \"./configs/config_maisi3d-rflow.json\"\n", + "else:\n", + " raise ValueError(f\"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}.\")\n", + "with open(model_def_path, \"r\") as f:\n", + " model_def = json.load(f)\n", + "include_body_region = model_def[\"include_body_region\"]\n", + "logger.info(f\"MAISI version is {maisi_version}, whether to use body_region is {include_body_region}\")" + ] + }, { "cell_type": "markdown", "id": "d8e29c23", @@ -95,7 +175,7 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, "id": "fc32a7fe", "metadata": {}, "outputs": [], @@ -117,7 +197,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 5, "id": "1b199078", "metadata": {}, "outputs": [ @@ -125,7 +205,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Generated simulated images.\n" + "[2025-03-14 16:14:22.301][ INFO](notebook) - Generated simulated images.\n" ] } ], @@ -154,7 +234,7 @@ }, { "cell_type": "markdown", - "id": "c2389853", + "id": "a059ddcf-8525-4241-9fe3-b661c4bdd336", "metadata": {}, "source": [ "### Set up directories and configurations\n", @@ -164,7 +244,7 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 6, "id": "6c7b434c", "metadata": {}, "outputs": [ @@ -172,15 +252,14 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:files and folders under work_dir: ['environment_maisi_diff_model.json', 'config_maisi.json', 'sim_dataroot', 'sim_datalist.json', 'models', 'embeddings', 'config_maisi_diff_model.json', 'predictions'].\n", - "INFO:notebook:number of GPUs: 1.\n" + "[2025-03-14 16:14:22.313][ INFO](notebook) - files and folders under work_dir: ['predictions', 'config_maisi.json', 'models', 'sim_dataroot', 'config_maisi_diff_model.json', 'embeddings', 'environment_maisi_diff_model.json', 'sim_datalist.json'].\n", + "[2025-03-14 16:14:22.314][ INFO](notebook) - number of GPUs: 1.\n" ] } ], "source": [ "env_config_path = \"./configs/environment_maisi_diff_model.json\"\n", "model_config_path = \"./configs/config_maisi_diff_model.json\"\n", - "model_def_path = \"./configs/config_maisi.json\"\n", "\n", "# Load environment configuration, model configuration and model definition\n", "with open(env_config_path, \"r\") as f:\n", @@ -189,9 +268,6 @@ "with open(model_config_path, \"r\") as f:\n", " model_config = json.load(f)\n", "\n", - "with open(model_def_path, \"r\") as f:\n", - " model_def = json.load(f)\n", - "\n", "env_config_out = copy.deepcopy(env_config)\n", "model_config_out = copy.deepcopy(model_config)\n", "model_def_out = copy.deepcopy(model_def)\n", @@ -229,7 +305,7 @@ " json.dump(model_config_out, f, sort_keys=True, indent=4)\n", "\n", "# Update model definition for demo\n", - "model_def_out[\"autoencoder_def\"][\"num_splits\"] = 4\n", + "model_def_out[\"autoencoder_def\"][\"num_splits\"] = 2\n", "model_def_filepath = os.path.join(work_dir, \"config_maisi.json\")\n", "with open(model_def_filepath, \"w\") as f:\n", " json.dump(model_def_out, f, sort_keys=True, indent=4)\n", @@ -244,7 +320,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 7, "id": "95ea6972", "metadata": {}, "outputs": [], @@ -304,7 +380,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "id": "f45ea863", "metadata": {}, "outputs": [ @@ -312,7 +388,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Creating training data...\n" + "[2025-03-14 16:14:22.326][ INFO](notebook) - Creating training data...\n" ] }, { @@ -320,8 +396,8 @@ "output_type": "stream", "text": [ "\n", - "INFO:creating training data:Using device cuda:0\n", - "INFO:creating training data:filenames_raw: ['tr_image_001.nii.gz', 'tr_image_002.nii.gz']\n", + "[2025-03-14 16:14:29.646][ INFO](creating training data) - Using device cuda:0\n", + "[2025-03-14 16:14:30.160][ INFO](creating training data) - filenames_raw: ['tr_image_001.nii.gz', 'tr_image_002.nii.gz']\n", "\n" ] } @@ -357,7 +433,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "id": "0221a658", "metadata": {}, "outputs": [ @@ -365,9 +441,11 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:data: {'dim': (64, 64, 32), 'spacing': [0.875, 0.875, 0.75], 'top_region_index': [0, 1, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}.\n", - "INFO:notebook:data: {'dim': (64, 64, 32), 'spacing': [0.875, 0.875, 0.75], 'top_region_index': [0, 1, 0, 0], 'bottom_region_index': [0, 0, 1, 0]}.\n", - "INFO:notebook:Completed creating .json files for all embedding files.\n" + "[2025-03-14 16:14:32.560][ INFO](notebook) - data: {'dim': (64, 64, 32), 'spacing': [0.875, 0.875, 0.75]}.\n", + "[2025-03-14 16:14:32.562][ INFO](notebook) - Save json file to ./temp_work_dir/./embeddings/tr_image_001_emb.nii.gz.json\n", + "[2025-03-14 16:14:32.563][ INFO](notebook) - data: {'dim': (64, 64, 32), 'spacing': [0.875, 0.875, 0.75]}.\n", + "[2025-03-14 16:14:32.564][ INFO](notebook) - Save json file to ./temp_work_dir/./embeddings/tr_image_002_emb.nii.gz.json\n", + "[2025-03-14 16:14:32.565][ INFO](notebook) - Completed creating .json files for all embedding files.\n" ] } ], @@ -396,15 +474,13 @@ " spacing = [float(_item) for _item in spacing]\n", "\n", " # Create the dictionary with the specified keys and values\n", - " # The region can be selected from one of four regions from top to bottom.\n", - " # [1,0,0,0] is the head and neck, [0,1,0,0] is the chest region, [0,0,1,0]\n", - " # is the abdomen region, and [0,0,0,1] is the lower body region.\n", - " data = {\n", - " \"dim\": dimensions,\n", - " \"spacing\": spacing,\n", - " \"top_region_index\": [0, 1, 0, 0], # chest region\n", - " \"bottom_region_index\": [0, 0, 1, 0], # abdomen region\n", - " }\n", + " data = {\"dim\": dimensions, \"spacing\": spacing}\n", + " if include_body_region:\n", + " # The region can be selected from one of four regions from top to bottom.\n", + " # [1,0,0,0] is the head and neck, [0,1,0,0] is the chest region, [0,0,1,0]\n", + " # is the abdomen region, and [0,0,0,1] is the lower body region.\n", + " data[\"top_region_index\"] = [0, 1, 0, 0] # chest region\n", + " data[\"bottom_region_index\"] = [0, 0, 1, 0] # abdomen region\n", " logger.info(f\"data: {data}.\")\n", "\n", " # Create the .json filename\n", @@ -413,6 +489,7 @@ " # Write the dictionary to the .json file\n", " with open(json_filename, \"w\") as json_file:\n", " json.dump(data, json_file, indent=4)\n", + " logger.info(f\"Save json file to {json_filename}\")\n", "\n", "\n", "folder_path = env_config_out[\"embedding_base_dir\"]\n", @@ -438,7 +515,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 10, "id": "ade6389d", "metadata": {}, "outputs": [ @@ -446,7 +523,7 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Training the model...\n" + "[2025-03-14 16:14:32.570][ INFO](notebook) - Training the model...\n" ] }, { @@ -454,26 +531,26 @@ "output_type": "stream", "text": [ "\n", - "INFO:training:Using cuda:0 of 1\n", - "INFO:training:[config] ckpt_folder -> ./temp_work_dir/./models.\n", - "INFO:training:[config] data_root -> ./temp_work_dir/./embeddings.\n", - "INFO:training:[config] data_list -> ./temp_work_dir/sim_datalist.json.\n", - "INFO:training:[config] lr -> 0.0001.\n", - "INFO:training:[config] num_epochs -> 2.\n", - "INFO:training:[config] num_train_timesteps -> 1000.\n", - "INFO:training:num_files_train: 2\n", - "INFO:training:Training from scratch.\n", - "INFO:training:Scaling factor set to 0.8903454542160034.\n", - "INFO:training:scale_factor -> 0.8903454542160034.\n", - "INFO:training:torch.set_float32_matmul_precision -> highest.\n", - "INFO:training:Epoch 1, lr 0.0001.\n", - "INFO:training:[2024-09-30 06:30:33] epoch 1, iter 1/2, loss: 0.7974, lr: 0.000100000000.\n", - "INFO:training:[2024-09-30 06:30:33] epoch 1, iter 2/2, loss: 0.7939, lr: 0.000056250000.\n", - "INFO:training:epoch 1 average loss: 0.7957.\n", - "INFO:training:Epoch 2, lr 2.5e-05.\n", - "INFO:training:[2024-09-30 06:30:35] epoch 2, iter 1/2, loss: 0.7902, lr: 0.000025000000.\n", - "INFO:training:[2024-09-30 06:30:35] epoch 2, iter 2/2, loss: 0.7889, lr: 0.000006250000.\n", - "INFO:training:epoch 2 average loss: 0.7895.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - Using cuda:0 of 1\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] ckpt_folder -> ./temp_work_dir/./models.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] data_root -> ./temp_work_dir/./embeddings.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] data_list -> ./temp_work_dir/sim_datalist.json.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] lr -> 0.0001.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] num_epochs -> 2.\n", + "[2025-03-14 16:14:39.869][ INFO](training) - [config] num_train_timesteps -> 1000.\n", + "[2025-03-14 16:14:41.316][ INFO](training) - Training from scratch.\n", + "[2025-03-14 16:14:41.337][ INFO](training) - num_files_train: 2\n", + "[2025-03-14 16:14:41.634][ INFO](training) - Scaling factor set to 1.159693956375122.\n", + "[2025-03-14 16:14:41.634][ INFO](training) - scale_factor -> 1.159693956375122.\n", + "[2025-03-14 16:14:41.637][ INFO](training) - torch.set_float32_matmul_precision -> highest.\n", + "[2025-03-14 16:14:41.637][ INFO](training) - Epoch 1, lr 0.0001.\n", + "[2025-03-14 16:14:42.627][ INFO](training) - [2025-03-14 16:14:42] epoch 1, iter 1/2, loss: 1.1344, lr: 0.000100000000.\n", + "[2025-03-14 16:14:42.739][ INFO](training) - [2025-03-14 16:14:42] epoch 1, iter 2/2, loss: 1.1275, lr: 0.000056250000.\n", + "[2025-03-14 16:14:42.783][ INFO](training) - epoch 1 average loss: 1.1310.\n", + "[2025-03-14 16:14:44.540][ INFO](training) - Epoch 2, lr 2.5e-05.\n", + "[2025-03-14 16:14:44.981][ INFO](training) - [2025-03-14 16:14:44] epoch 2, iter 1/2, loss: 1.1254, lr: 0.000025000000.\n", + "[2025-03-14 16:14:45.106][ INFO](training) - [2025-03-14 16:14:45] epoch 2, iter 2/2, loss: 1.1201, lr: 0.000006250000.\n", + "[2025-03-14 16:14:45.177][ INFO](training) - epoch 2 average loss: 1.1227.\n", "\n" ] } @@ -509,7 +586,7 @@ }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 11, "id": "1626526d", "metadata": {}, "outputs": [ @@ -517,8 +594,8 @@ "name": "stderr", "output_type": "stream", "text": [ - "INFO:notebook:Running inference...\n", - "INFO:notebook:Completed all steps.\n" + "[2025-03-14 16:14:49.136][ INFO](notebook) - Running inference...\n", + "[2025-03-14 16:15:02.647][ INFO](notebook) - Completed all steps.\n" ] }, { @@ -526,24 +603,24 @@ "output_type": "stream", "text": [ "\n", - "INFO:inference:Using cuda:0 of 1 with random seed: 93612\n", - "INFO:inference:[config] ckpt_filepath -> ./temp_work_dir/./models/diff_unet_ckpt.pt.\n", - "INFO:inference:[config] random_seed -> 93612.\n", - "INFO:inference:[config] output_prefix -> unet_3d.\n", - "INFO:inference:[config] output_size -> (256, 256, 128).\n", - "INFO:inference:[config] out_spacing -> (1.0, 1.0, 0.75).\n", - "INFO:root:`controllable_anatomy_size` is not provided.\n", - "INFO:inference:checkpoints ./temp_work_dir/./models/diff_unet_ckpt.pt loaded.\n", - "INFO:inference:scale_factor -> 0.8903454542160034.\n", - "INFO:inference:num_downsample_level -> 4, divisor -> 4.\n", - "INFO:inference:noise: cuda:0, torch.float32, \n", + "[2025-03-14 16:14:56.275][ INFO](inference) - Using cuda:0 of 1 with random seed: 59473\n", + "[2025-03-14 16:14:56.275][ INFO](inference) - [config] ckpt_filepath -> ./temp_work_dir/./models/diff_unet_ckpt.pt.\n", + "[2025-03-14 16:14:56.275][ INFO](inference) - [config] random_seed -> 59473.\n", + "[2025-03-14 16:14:56.275][ INFO](inference) - [config] output_prefix -> unet_3d.\n", + "[2025-03-14 16:14:56.275][ INFO](inference) - [config] output_size -> (256, 256, 128).\n", + "[2025-03-14 16:14:56.275][ INFO](inference) - [config] out_spacing -> (1.0, 1.0, 0.75).\n", + "[2025-03-14 16:14:56.275][ INFO](root) - `controllable_anatomy_size` is not provided.\n", + "[2025-03-14 16:14:58.525][ INFO](inference) - checkpoints ./temp_work_dir/./models/diff_unet_ckpt.pt loaded.\n", + "[2025-03-14 16:14:58.527][ INFO](inference) - scale_factor -> 1.159693956375122.\n", + "[2025-03-14 16:14:58.528][ INFO](inference) - num_downsample_level -> 4, divisor -> 4.\n", + "[2025-03-14 16:14:58.536][ INFO](inference) - noise: cuda:0, torch.float32, \n", "\n", - " 0%| | 0/10 [00:00 Tensor: - """ - Initialize tumor mask for tumor augmentation. +def erode3d(input_tensor, erosion=3): + # Define the structuring element + erosion = ensure_tuple_rep(erosion, 3) + structuring_element = torch.ones(1, 1, erosion[0], erosion[1], erosion[2]).to(input_tensor.device) - Args: - volume: input 3D multi-label mask, [1,H,W,D] torch tensor. - tumor_label: tumor label in whole_mask, list of int. - - Return: - tumor_mask_, initialized tumor mask, [1,H,W,D] torch tensor. - """ - tumor_mask_ = torch.zeros_like(volume, dtype=torch.uint8) - for idx, label in enumerate(tumor_label): - tumor_mask_[volume == label] = idx + 1 - return tumor_mask_ + # Pad the input tensor to handle border pixels + input_padded = F.pad( + input_tensor.float().unsqueeze(0).unsqueeze(0), + (erosion[0] // 2, erosion[0] // 2, erosion[1] // 2, erosion[1] // 2, erosion[2] // 2, erosion[2] // 2), + mode="constant", + value=1.0, + ) + # Apply erosion operation + output = F.conv3d(input_padded, structuring_element, padding=0) -def finalize_tumor_mask(augmented_mask: Tensor, organ_mask: Tensor, threshold_tumor_size: float): - """ - Try to generate the final tumor mask by combining the augmented tumor mask and organ mask. - Need to make sure tumor is inside of organ and is larger than threshold_tumor_size. + # Set output values based on the minimum value within the structuring element + output = torch.where(output == torch.sum(structuring_element), 1.0, 0.0) - Args: - augmented_mask: input 3D binary tumor mask, [1,H,W,D] torch tensor. - organ_mask: input 3D binary organ mask, [1,H,W,D] torch tensor. - threshold_tumor_size: threshold tumor size, float + return output.squeeze(0).squeeze(0) - Return: - tumor_mask, [H,W,D] torch tensor; or None if the size did not qualify - """ - tumor_mask = augmented_mask * organ_mask - if torch.sum(tumor_mask) >= threshold_tumor_size: - tumor_mask = dilate_one_img(tumor_mask.squeeze(0), filter_size=5, pad_value=1.0) - tumor_mask = erode_one_img(tumor_mask, filter_size=5, pad_value=1.0).unsqueeze(0).to(torch.uint8) - return tumor_mask - else: - return None +def dilate3d(input_tensor, erosion=3): + # Define the structuring element + erosion = ensure_tuple_rep(erosion, 3) + structuring_element = torch.ones(1, 1, erosion[0], erosion[1], erosion[2]).to(input_tensor.device) -def augmentation_bone_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Bone tumor augmentation. + # Pad the input tensor to handle border pixels + input_padded = F.pad( + input_tensor.float().unsqueeze(0).unsqueeze(0), + (erosion[0] // 2, erosion[0] // 2, erosion[1] // 2, erosion[1] // 2, erosion[2] // 2, erosion[2] // 2), + mode="constant", + value=1.0, + ) - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. - If not defined, will use (H,W,D). If some components are non-positive values, - the transform will use the corresponding components of whole_mask size. - For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) - if the third spatial dimension size of whole_mask is 64. + # Apply erosion operation + output = F.conv3d(input_padded, structuring_element, padding=0) - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. + # Set output values based on the minimum value within the structuring element + output = torch.where(output > 0, 1.0, 0.0) - Example: + return output.squeeze(0).squeeze(0) - .. code-block:: python - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=127 - whole_mask[0,0, 97:103, 97:103, 97:103]=128 - augmented_whole_mask = augmentation_bone_tumor(whole_mask) - """ - # Initialize binary tumor mask - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - tumor_label = [128] - tumor_mask_ = initialize_tumor_mask(volume, tumor_label) +def augmentation_tumor_bone(pt_nda, output_size, random_seed): + volume = pt_nda.squeeze(0) + real_l_volume_ = torch.zeros_like(volume) + real_l_volume_[volume == 128] = 1 + real_l_volume_ = real_l_volume_.to(torch.uint8) - # Define augmentation transform elastic = RandAffine( mode="nearest", prob=1.0, @@ -105,78 +74,52 @@ def augmentation_bone_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, in scale_range=(0.15, 0.15, 0), padding_mode="zeros", ) + elastic.set_random_state(seed=random_seed) - tumor_size = torch.sum((tumor_mask_ > 0).float()) + tumor_szie = torch.sum((real_l_volume_ > 0).float()) ########################### # remove pred in pseudo_label in real lesion region - volume[tumor_mask_ > 0] = 200 + volume[real_l_volume_ > 0] = 200 ########################### - if tumor_size > 0: + if tumor_szie > 0: # get organ mask organ_mask = ( torch.logical_and(33 <= volume, volume <= 56).float() + torch.logical_and(63 <= volume, volume <= 97).float() + (volume == 127).float() + (volume == 114).float() - + tumor_mask_ + + real_l_volume_ ) organ_mask = (organ_mask > 0).float() - - # augment mask - count = 0 + cnt = 0 while True: - threshold = 0.8 if count < 40 else 0.75 - tumor_mask = tumor_mask_ - # apply random augmentation - augmented_mask = elastic(tumor_mask > 0, spatial_size=spatial_size).as_tensor() - # generate final tumor mask - count += 1 - tumor_mask = finalize_tumor_mask(augmented_mask, organ_mask, tumor_size * threshold) - if tumor_mask is not None: + threshold = 0.8 if cnt < 40 else 0.75 + real_l_volume = real_l_volume_ + # random distor mask + distored_mask = elastic((real_l_volume > 0).cuda(), spatial_size=tuple(output_size)).as_tensor() + real_l_volume = distored_mask * organ_mask + cnt += 1 + print(torch.sum(real_l_volume), "|", tumor_szie * threshold) + if torch.sum(real_l_volume) >= tumor_szie * threshold: + real_l_volume = dilate3d(real_l_volume.squeeze(0), erosion=5) + real_l_volume = erode3d(real_l_volume, erosion=5).unsqueeze(0).to(torch.uint8) break - if count > MAX_COUNT: - raise ValueError("Please check if bone lesion is inside bone.") else: - tumor_mask = tumor_mask_ - - # update the new tumor mask - volume[tumor_mask == 1] = tumor_label[0] - - return volume.unsqueeze(0).to(device) - - -def augmentation_liver_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Bone liver augmentation. + real_l_volume = real_l_volume_ - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. - If not defined, will use (H,W,D). If some components are non-positive values, - the transform will use the corresponding components of whole_mask size. - For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) - if the third spatial dimension size of whole_mask is 64. + volume[real_l_volume == 1] = 128 - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. + pt_nda = volume.unsqueeze(0) + return pt_nda - Example: - .. code-block:: python +def augmentation_tumor_liver(pt_nda, output_size, random_seed): + volume = pt_nda.squeeze(0) + real_l_volume_ = torch.zeros_like(volume) + real_l_volume_[volume == 1] = 1 + real_l_volume_[volume == 26] = 2 + real_l_volume_ = real_l_volume_.to(torch.uint8) - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=1 - whole_mask[0,0, 97:103, 97:103, 97:103]=26 - augmented_whole_mask = augment_liver_tumor(whole_mask) - """ - # Initialize binary tumor mask - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - tumor_label = [1, 26] - tumor_mask_ = initialize_tumor_mask(volume, tumor_label) - - # Define augmentation transform elastic = Rand3DElastic( mode="nearest", prob=1.0, @@ -187,74 +130,45 @@ def augmentation_liver_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, i scale_range=(0.2, 0.2, 0.2), padding_mode="zeros", ) + elastic.set_random_state(seed=random_seed) - tumor_size = torch.sum(tumor_mask_ == 2) + tumor_szie = torch.sum(real_l_volume_ == 2) ########################### # remove pred organ labels volume[volume == 1] = 0 volume[volume == 26] = 0 # before move tumor maks, full the original location by organ labels - volume[tumor_mask_ == 1] = 1 - volume[tumor_mask_ == 2] = 1 + volume[real_l_volume_ == 1] = 1 + volume[real_l_volume_ == 2] = 1 ########################### - if tumor_size > 0: - count = 0 + while True: + real_l_volume = real_l_volume_ + # random distor mask + real_l_volume = elastic((real_l_volume == 2).cuda(), spatial_size=tuple(output_size)).as_tensor() # get organ mask - organ_mask = (tumor_mask_ == 1).float() + (tumor_mask_ == 2).float() - organ_mask = dilate_one_img(organ_mask.squeeze(0), filter_size=5, pad_value=1.0) - organ_mask = erode_one_img(organ_mask, filter_size=5, pad_value=1.0).unsqueeze(0) - while True: - tumor_mask = tumor_mask_ - # apply random augmentation - augmented_mask = elastic((tumor_mask == 2), spatial_size=spatial_size).as_tensor() - - # generate final tumor mask - count += 1 - tumor_mask = finalize_tumor_mask(augmented_mask, organ_mask, tumor_size * 0.80) - if tumor_mask is not None: - break - if count > MAX_COUNT: - raise ValueError("Please check if liver tumor is inside liver.") - else: - tumor_mask = tumor_mask_ - - volume[tumor_mask == 1] = 26 - - return volume.unsqueeze(0).to(device) - + organ_mask = (real_l_volume_ == 1).float() + (real_l_volume_ == 2).float() -def augmentation_lung_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Lung tumor augmentation. + organ_mask = dilate3d(organ_mask.squeeze(0), erosion=5) + organ_mask = erode3d(organ_mask, erosion=5).unsqueeze(0) + real_l_volume = real_l_volume * organ_mask + print(torch.sum(real_l_volume), "|", tumor_szie * 0.80) + if torch.sum(real_l_volume) >= tumor_szie * 0.80: + real_l_volume = dilate3d(real_l_volume.squeeze(0), erosion=5) + real_l_volume = erode3d(real_l_volume, erosion=5).unsqueeze(0) + break - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. - If not defined, will use (H,W,D). If some components are non-positive values, - the transform will use the corresponding components of whole_mask size. - For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) - if the third spatial dimension size of whole_mask is 64. + volume[real_l_volume == 1] = 26 - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. + pt_nda = volume.unsqueeze(0) + return pt_nda - Example: - .. code-block:: python +def augmentation_tumor_lung(pt_nda, output_size, random_seed): + volume = pt_nda.squeeze(0) + real_l_volume_ = torch.zeros_like(volume) + real_l_volume_[volume == 23] = 1 + real_l_volume_ = real_l_volume_.to(torch.uint8) - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=28 - whole_mask[0,0, 97:103, 97:103, 97:103]=23 - augmented_whole_mask = augmentation_lung_tumor(whole_mask) - """ - # Initialize binary tumor mask - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - tumor_label = [23] - tumor_mask_ = initialize_tumor_mask(volume, tumor_label) - - # Define augmentation transform elastic = Rand3DElastic( mode="nearest", prob=1.0, @@ -265,87 +179,61 @@ def augmentation_lung_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, in scale_range=(0.15, 0.15, 0.15), padding_mode="zeros", ) + elastic.set_random_state(seed=random_seed) - tumor_size = torch.sum(tumor_mask_) + tumor_szie = torch.sum(real_l_volume_) # before move lung tumor maks, full the original location by lung labels - new_tumor_mask_ = dilate_one_img(tumor_mask_.squeeze(0), filter_size=3, pad_value=1.0) - new_tumor_mask_ = new_tumor_mask_.unsqueeze(0) - new_tumor_mask_[tumor_mask_ > 0] = 0 - new_tumor_mask_[volume < 28] = 0 - new_tumor_mask_[volume > 32] = 0 - tmp = volume[(volume * new_tumor_mask_).nonzero(as_tuple=True)].view(-1) + new_real_l_volume_ = dilate3d(real_l_volume_.squeeze(0), erosion=3) + new_real_l_volume_ = new_real_l_volume_.unsqueeze(0) + new_real_l_volume_[real_l_volume_ > 0] = 0 + new_real_l_volume_[volume < 28] = 0 + new_real_l_volume_[volume > 32] = 0 + tmp = volume[(volume * new_real_l_volume_).nonzero(as_tuple=True)].view(-1) mode = torch.mode(tmp, 0)[0].item() + print(mode) assert 28 <= mode <= 32 - volume[tumor_mask_.bool()] = mode + volume[real_l_volume_.bool()] = mode ########################### - - if tumor_size > 0: - count = 0 - # get lung mask v2 (133 order) - organ_mask = ( - (volume == 28).float() - + (volume == 29).float() - + (volume == 30).float() - + (volume == 31).float() - + (volume == 32).float() - ) - organ_mask = dilate_one_img(organ_mask.squeeze(0), filter_size=5, pad_value=1.0) - organ_mask = erode_one_img(organ_mask, filter_size=5, pad_value=1.0).unsqueeze(0) - + if tumor_szie > 0: # aug while True: - tumor_mask = tumor_mask_ - # apply random augmentation - augmented_mask = elastic(tumor_mask, spatial_size=spatial_size).as_tensor() - - # generate final tumor mask - count += 1 - tumor_mask = finalize_tumor_mask(augmented_mask, organ_mask, tumor_size * 0.85) - if tumor_mask is not None: + real_l_volume = real_l_volume_ + # random distor mask + real_l_volume = elastic(real_l_volume, spatial_size=tuple(output_size)).as_tensor() + # get lung mask v2 (133 order) + lung_mask = ( + (volume == 28).float() + + (volume == 29).float() + + (volume == 30).float() + + (volume == 31).float() + + (volume == 32).float() + ) + + lung_mask = dilate3d(lung_mask.squeeze(0), erosion=5) + lung_mask = erode3d(lung_mask, erosion=5).unsqueeze(0) + real_l_volume = real_l_volume * lung_mask + print(torch.sum(real_l_volume), "|", tumor_szie * 0.85) + if torch.sum(real_l_volume) >= tumor_szie * 0.85: + real_l_volume = dilate3d(real_l_volume.squeeze(0), erosion=5) + real_l_volume = erode3d(real_l_volume, erosion=5).unsqueeze(0).to(torch.uint8) break - if count > MAX_COUNT: - raise ValueError("Please check if lung tumor is inside lung.") else: - tumor_mask = tumor_mask_ - - volume[tumor_mask == 1] = tumor_label[0] + real_l_volume = real_l_volume_ - return volume.unsqueeze(0).to(device) + volume[real_l_volume == 1] = 23 + pt_nda = volume.unsqueeze(0) + return pt_nda -def augmentation_pancreas_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Pancreas tumor augmentation. - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. - If not defined, will use (H,W,D). If some components are non-positive values, - the transform will use the corresponding components of whole_mask size. - For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) - if the third spatial dimension size of whole_mask is 64. +def augmentation_tumor_pancreas(pt_nda, output_size, random_seed): + volume = pt_nda.squeeze(0) + real_l_volume_ = torch.zeros_like(volume) + real_l_volume_[volume == 4] = 1 + real_l_volume_[volume == 24] = 2 + real_l_volume_ = real_l_volume_.to(torch.uint8) - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. - - Example: - - .. code-block:: python - - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=24 - whole_mask[0,0, 97:103, 97:103, 97:103]=4 - augmented_whole_mask = augmentation_pancreas_tumor(whole_mask) - """ - # Initialize binary tumor mask - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - tumor_label = [4, 24] - tumor_mask_ = initialize_tumor_mask(volume, tumor_label) - - # Define augmentation transform elastic = Rand3DElastic( mode="nearest", prob=1.0, @@ -356,74 +244,45 @@ def augmentation_pancreas_tumor(whole_mask: Tensor, spatial_size: tuple[int, int scale_range=(0.1, 0.1, 0.1), padding_mode="zeros", ) + elastic.set_random_state(seed=random_seed) - tumor_size = torch.sum(tumor_mask_ == 2) + tumor_szie = torch.sum(real_l_volume_ == 2) ########################### # remove pred organ labels volume[volume == 24] = 0 volume[volume == 4] = 0 # before move tumor maks, full the original location by organ labels - volume[tumor_mask_ == 1] = 4 - volume[tumor_mask_ == 2] = 4 + volume[real_l_volume_ == 1] = 4 + volume[real_l_volume_ == 2] = 4 ########################### - if tumor_size > 0: - count = 0 + while True: + real_l_volume = real_l_volume_ + # random distor mask + real_l_volume = elastic((real_l_volume == 2).cuda(), spatial_size=tuple(output_size)).as_tensor() # get organ mask - organ_mask = (tumor_mask_ == 1).float() + (tumor_mask_ == 2).float() - organ_mask = dilate_one_img(organ_mask.squeeze(0), filter_size=5, pad_value=1.0) - organ_mask = erode_one_img(organ_mask, filter_size=5, pad_value=1.0).unsqueeze(0) - while True: - tumor_mask = tumor_mask_ - # apply random augmentation - augmented_mask = elastic((tumor_mask == 2), spatial_size=spatial_size).as_tensor() - - # generate final tumor mask - count += 1 - tumor_mask = finalize_tumor_mask(augmented_mask, organ_mask, tumor_size * 0.80) - if tumor_mask is not None: - break - if count > MAX_COUNT: - raise ValueError("Please check if pancreas tumor is inside pancreas.") - else: - tumor_mask = tumor_mask_ + organ_mask = (real_l_volume_ == 1).float() + (real_l_volume_ == 2).float() - volume[tumor_mask == 1] = 24 + organ_mask = dilate3d(organ_mask.squeeze(0), erosion=5) + organ_mask = erode3d(organ_mask, erosion=5).unsqueeze(0) + real_l_volume = real_l_volume * organ_mask + print(torch.sum(real_l_volume), "|", tumor_szie * 0.80) + if torch.sum(real_l_volume) >= tumor_szie * 0.80: + real_l_volume = dilate3d(real_l_volume.squeeze(0), erosion=5) + real_l_volume = erode3d(real_l_volume, erosion=5).unsqueeze(0) + break - return volume.unsqueeze(0).to(device) + volume[real_l_volume == 1] = 24 + pt_nda = volume.unsqueeze(0) + return pt_nda -def augmentation_colon_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Colon tumor augmentation. - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. - If not defined, will use (H,W,D). If some components are non-positive values, - the transform will use the corresponding components of whole_mask size. - For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) - if the third spatial dimension size of whole_mask is 64. +def augmentation_tumor_colon(pt_nda, output_size, random_seed): + volume = pt_nda.squeeze(0) + real_l_volume_ = torch.zeros_like(volume) + real_l_volume_[volume == 27] = 1 + real_l_volume_ = real_l_volume_.to(torch.uint8) - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. - - Example: - - .. code-block:: python - - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=62 - whole_mask[0,0, 97:103, 97:103, 97:103]=27 - augmented_whole_mask = augmentation_colon_tumor(whole_mask) - """ - # Initialize binary tumor mask - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - tumor_label = [27] - tumor_mask_ = initialize_tumor_mask(volume, tumor_label) - - # Define augmentation transform elastic = Rand3DElastic( mode="nearest", prob=1.0, @@ -434,125 +293,81 @@ def augmentation_colon_tumor(whole_mask: Tensor, spatial_size: tuple[int, int, i scale_range=(0.1, 0.1, 0.1), padding_mode="zeros", ) + elastic.set_random_state(seed=random_seed) - tumor_size = torch.sum(tumor_mask_) + tumor_szie = torch.sum(real_l_volume_) ########################### # before move tumor maks, full the original location by organ labels - volume[tumor_mask_.bool()] = 62 + volume[real_l_volume_.bool()] = 62 ########################### - if tumor_size > 0: + if tumor_szie > 0: # get organ mask organ_mask = (volume == 62).float() - organ_mask = dilate_one_img(organ_mask.squeeze(0), filter_size=5, pad_value=1.0) - organ_mask = erode_one_img(organ_mask, filter_size=5, pad_value=1.0).unsqueeze(0) - - count = 0 + organ_mask = dilate3d(organ_mask.squeeze(0), erosion=5) + organ_mask = erode3d(organ_mask, erosion=5).unsqueeze(0) + # cnt = 0 + cnt = 0 while True: threshold = 0.8 - tumor_mask = tumor_mask_ - if count < 20: - # apply random augmentation - augmented_mask = elastic((tumor_mask == 1), spatial_size=spatial_size).as_tensor() - tumor_mask = augmented_mask * organ_mask - elif 20 <= count < 40: + real_l_volume = real_l_volume_ + if cnt < 20: + # random distor mask + distored_mask = elastic((real_l_volume == 1).cuda(), spatial_size=tuple(output_size)).as_tensor() + real_l_volume = distored_mask * organ_mask + elif 20 <= cnt < 40: threshold = 0.75 else: break - # generate final tumor mask - count += 1 - tumor_mask = finalize_tumor_mask(tumor_mask, organ_mask, tumor_size * threshold) - if tumor_mask is not None: + real_l_volume = real_l_volume * organ_mask + print(torch.sum(real_l_volume), "|", tumor_szie * threshold) + cnt += 1 + if torch.sum(real_l_volume) >= tumor_szie * threshold: + real_l_volume = dilate3d(real_l_volume.squeeze(0), erosion=5) + real_l_volume = erode3d(real_l_volume, erosion=5).unsqueeze(0).to(torch.uint8) break - if count > MAX_COUNT: - raise ValueError("Please check if colon tumor is inside colon.") else: - tumor_mask = tumor_mask_ - - volume[tumor_mask == 1] = tumor_label[0] + real_l_volume = real_l_volume_ + # break + volume[real_l_volume == 1] = 27 - return volume.unsqueeze(0).to(device) + pt_nda = volume.unsqueeze(0) + return pt_nda -def augmentation_body(whole_mask: Tensor) -> Tensor: - """ - Whole body mask augmentation. +def augmentation_body(pt_nda, random_seed): + volume = pt_nda.squeeze(0) - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. + zoom = RandZoom(min_zoom=0.99, max_zoom=1.01, mode="nearest", align_corners=None, prob=1.0) + zoom.set_random_state(seed=random_seed) - Return: - augmented mask, with same shape and data type as whole_mask. - - Example: - - .. code-block:: python - - # define a multi-label mask - whole_mask = torch.zeros([1,1,128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=127 - whole_mask[0,0, 97:103, 97:103, 97:103]=128 - augmented_whole_mask = augmentation_body(whole_mask) - """ - device = whole_mask.device - volume = whole_mask.squeeze(0).cuda() if not whole_mask.is_cuda else whole_mask.squeeze(0) - - # Define augmentation transform - zoom = RandZoom( - min_zoom=0.99, - max_zoom=1.01, - mode="nearest", - align_corners=None, - prob=1.0, - ) - # apply random augmentation volume = zoom(volume) - return volume.unsqueeze(0).to(device) - - -def augmentation(whole_mask: Tensor, spatial_size: tuple[int, int, int] | int | None = None) -> Tensor: - """ - Tumor or whole body mask augmentation. If tumor exist, augment tumor mask; if not, augment whole body mask - - Args: - whole_mask: input 3D multi-label mask, [1,1,H,W,D] torch tensor. - spatial_size: output image spatial size, used in random transform. If not defined, will use (H,W,D). If some components are non-positive values, the transform will use the corresponding components of whole_mask size. For example, spatial_size=(128, 128, -1) will be adapted to (128, 128, 64) if the third spatial dimension size of whole_mask is 64. - - Return: - augmented mask, with shape of spatial_size and data type as whole_mask. - - Example: + pt_nda = volume.unsqueeze(0) + return pt_nda - .. code-block:: python - # define a multi-label mask - whole_mask = torch.zeros([1,1, 128,128,128]) - whole_mask[0,0, 90:110, 90:110, 90:110]=127 - whole_mask[0,0, 97:103, 97:103, 97:103]=128 - augmented_whole_mask = augmentation(whole_mask) - """ - label_list = torch.unique(whole_mask) +def augmentation(pt_nda, output_size, random_seed): + label_list = torch.unique(pt_nda) label_list = list(label_list.cpu().numpy()) - # Note that we only augment one type of tumor. if 128 in label_list: print("augmenting bone lesion/tumor") - whole_mask = augmentation_bone_tumor(whole_mask, spatial_size) + pt_nda = augmentation_tumor_bone(pt_nda, output_size, random_seed) elif 26 in label_list: print("augmenting liver tumor") - whole_mask = augmentation_liver_tumor(whole_mask, spatial_size) + pt_nda = augmentation_tumor_liver(pt_nda, output_size, random_seed) elif 23 in label_list: print("augmenting lung tumor") - whole_mask = augmentation_lung_tumor(whole_mask, spatial_size) + pt_nda = augmentation_tumor_lung(pt_nda, output_size, random_seed) elif 24 in label_list: print("augmenting pancreas tumor") - whole_mask = augmentation_pancreas_tumor(whole_mask, spatial_size) + pt_nda = augmentation_tumor_pancreas(pt_nda, output_size, random_seed) elif 27 in label_list: print("augmenting colon tumor") - whole_mask = augmentation_colon_tumor(whole_mask, spatial_size) + pt_nda = augmentation_tumor_colon(pt_nda, output_size, random_seed) else: print("augmenting body") - whole_mask = augmentation_body(whole_mask) + pt_nda = augmentation_body(pt_nda, random_seed) - return whole_mask + return pt_nda diff --git a/generation/maisi/scripts/diff_model_create_training_data.py b/generation/maisi/scripts/diff_model_create_training_data.py index ca44b43cc7..177dfa34cf 100644 --- a/generation/maisi/scripts/diff_model_create_training_data.py +++ b/generation/maisi/scripts/diff_model_create_training_data.py @@ -17,12 +17,11 @@ import os from pathlib import Path +import monai import nibabel as nib import numpy as np import torch import torch.distributed as dist - -import monai from monai.transforms import Compose from monai.utils import set_determinism diff --git a/generation/maisi/scripts/diff_model_infer.py b/generation/maisi/scripts/diff_model_infer.py index 9ba837328c..8b01e1cc96 100644 --- a/generation/maisi/scripts/diff_model_infer.py +++ b/generation/maisi/scripts/diff_model_infer.py @@ -21,14 +21,15 @@ import numpy as np import torch import torch.distributed as dist -from tqdm import tqdm - from monai.inferers import sliding_window_inference +from monai.inferers.inferer import SlidingWindowInferer +from monai.networks.schedulers import RFlowScheduler from monai.utils import set_determinism +from tqdm import tqdm from .diff_model_setting import initialize_distributed, load_config, setup_logging from .sample import ReconModel, check_input -from .utils import define_instance +from .utils import define_instance, dynamic_infer def set_random_seed(seed: int) -> int: @@ -94,8 +95,11 @@ def prepare_tensors(args: argparse.Namespace, device: torch.device) -> tuple: top_region_index_tensor = torch.from_numpy(top_region_index_tensor[np.newaxis, :]).half().to(device) bottom_region_index_tensor = torch.from_numpy(bottom_region_index_tensor[np.newaxis, :]).half().to(device) spacing_tensor = torch.from_numpy(spacing_tensor[np.newaxis, :]).half().to(device) + modality_tensor = args.diffusion_unet_inference["modality"] * torch.ones( + (len(spacing_tensor)), dtype=torch.long + ).to(device) - return top_region_index_tensor, bottom_region_index_tensor, spacing_tensor + return top_region_index_tensor, bottom_region_index_tensor, spacing_tensor, modality_tensor def run_inference( @@ -107,6 +111,7 @@ def run_inference( top_region_index_tensor: torch.Tensor, bottom_region_index_tensor: torch.Tensor, spacing_tensor: torch.Tensor, + modality_tensor: torch.Tensor, output_size: tuple, divisor: int, logger: logging.Logger, @@ -123,6 +128,7 @@ def run_inference( top_region_index_tensor (torch.Tensor): Top region index tensor. bottom_region_index_tensor (torch.Tensor): Bottom region index tensor. spacing_tensor (torch.Tensor): Spacing tensor. + modality_tensor (torch.Tensor): Modality tensor. output_size (tuple): Output size of the synthetic image. divisor (int): Divisor for downsample level. logger (logging.Logger): Logger for logging information. @@ -130,6 +136,9 @@ def run_inference( Returns: np.ndarray: Generated synthetic image data. """ + include_body_region = unet.include_top_region_index_input + include_modality = unet.num_class_embeds is not None + noise = torch.randn( ( 1, @@ -144,38 +153,64 @@ def run_inference( image = noise noise_scheduler = define_instance(args, "noise_scheduler") - noise_scheduler.set_timesteps(num_inference_steps=args.diffusion_unet_inference["num_inference_steps"]) + if isinstance(noise_scheduler, RFlowScheduler): + noise_scheduler.set_timesteps( + num_inference_steps=args.diffusion_unet_inference["num_inference_steps"], + input_img_size_numel=torch.prod(torch.tensor(noise.shape[2:])), + ) + else: + noise_scheduler.set_timesteps(num_inference_steps=args.diffusion_unet_inference["num_inference_steps"]) recon_model = ReconModel(autoencoder=autoencoder, scale_factor=scale_factor).to(device) autoencoder.eval() unet.eval() + all_timesteps = noise_scheduler.timesteps + all_next_timesteps = torch.cat((all_timesteps[1:], torch.tensor([0], dtype=all_timesteps.dtype))) + progress_bar = tqdm( + zip(all_timesteps, all_next_timesteps), + total=min(len(all_timesteps), len(all_next_timesteps)), + ) with torch.amp.autocast("cuda", enabled=True): - for t in tqdm(noise_scheduler.timesteps, ncols=110): - model_output = unet( - x=image, - timesteps=torch.Tensor((t,)).to(device), - top_region_index_tensor=top_region_index_tensor, - bottom_region_index_tensor=bottom_region_index_tensor, - spacing_tensor=spacing_tensor, - ) - image, _ = noise_scheduler.step(model_output, t, image) - - synthetic_images = sliding_window_inference( - inputs=image, - roi_size=( - min(output_size[0] // divisor // 4 * 3, 96), - min(output_size[1] // divisor // 4 * 3, 96), - min(output_size[2] // divisor // 4 * 3, 96), - ), + for t, next_t in progress_bar: + # Create a dictionary to store the inputs + unet_inputs = { + "x": image, + "timesteps": torch.Tensor((t,)).to(device), + "spacing_tensor": spacing_tensor, + } + + # Add extra arguments if include_body_region is True + if include_body_region: + unet_inputs.update( + { + "top_region_index_tensor": top_region_index_tensor, + "bottom_region_index_tensor": bottom_region_index_tensor, + } + ) + + if include_modality: + unet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + model_output = unet(**unet_inputs) + if not isinstance(noise_scheduler, RFlowScheduler): + image, _ = noise_scheduler.step(model_output, t, image) # type: ignore + else: + image, _ = noise_scheduler.step(model_output, t, image, next_t) # type: ignore + + inferer = SlidingWindowInferer( + roi_size=[80, 80, 80], sw_batch_size=1, - predictor=recon_model, + progress=True, mode="gaussian", - overlap=2.0 / 3.0, + overlap=0.4, sw_device=device, device=device, ) - + synthetic_images = dynamic_infer(inferer, recon_model, image) data = synthetic_images.squeeze().cpu().detach().numpy() a_min, a_max, b_min, b_max = -1000, 1000, 0, 1 data = (data - b_min) / (b_max - b_min) * (a_max - a_min) + a_min @@ -256,7 +291,7 @@ def diff_model_infer(env_config_path: str, model_config_path: str, model_def_pat divisor = 2 ** (num_downsample_level - 2) logger.info(f"num_downsample_level -> {num_downsample_level}, divisor -> {divisor}.") - top_region_index_tensor, bottom_region_index_tensor, spacing_tensor = prepare_tensors(args, device) + top_region_index_tensor, bottom_region_index_tensor, spacing_tensor, modality_tensor = prepare_tensors(args, device) data = run_inference( args, device, @@ -266,6 +301,7 @@ def diff_model_infer(env_config_path: str, model_config_path: str, model_def_pat top_region_index_tensor, bottom_region_index_tensor, spacing_tensor, + modality_tensor, output_size, divisor, logger, diff --git a/generation/maisi/scripts/diff_model_setting.py b/generation/maisi/scripts/diff_model_setting.py index 6ba4688867..3118b56d07 100644 --- a/generation/maisi/scripts/diff_model_setting.py +++ b/generation/maisi/scripts/diff_model_setting.py @@ -17,7 +17,6 @@ import torch import torch.distributed as dist - from monai.utils import RankFilter diff --git a/generation/maisi/scripts/diff_model_train.py b/generation/maisi/scripts/diff_model_train.py index 2309c8a4ee..c616b89c37 100644 --- a/generation/maisi/scripts/diff_model_train.py +++ b/generation/maisi/scripts/diff_model_train.py @@ -18,15 +18,16 @@ from datetime import datetime from pathlib import Path +import monai import torch import torch.distributed as dist -from torch.amp import GradScaler, autocast -from torch.nn.parallel import DistributedDataParallel - -import monai from monai.data import DataLoader, partition_dataset +from monai.networks.schedulers import RFlowScheduler +from monai.networks.schedulers.ddpm import DDPMPredictionType from monai.transforms import Compose from monai.utils import first +from torch.amp import GradScaler, autocast +from torch.nn.parallel import DistributedDataParallel from .diff_model_setting import initialize_distributed, load_config, setup_logging from .utils import define_instance @@ -49,7 +50,12 @@ def load_filenames(data_list_path: str) -> list: def prepare_data( - train_files: list, device: torch.device, cache_rate: float, num_workers: int = 2, batch_size: int = 1 + train_files: list, + device: torch.device, + cache_rate: float, + num_workers: int = 2, + batch_size: int = 1, + include_body_region: bool = False, ) -> DataLoader: """ Prepare training data. @@ -60,6 +66,7 @@ def prepare_data( cache_rate (float): Cache rate for dataset. num_workers (int): Number of workers for data loading. batch_size (int): Mini-batch size. + include_body_region (bool): Whether to include body region in data Returns: DataLoader: Data loader for training. @@ -69,22 +76,24 @@ def _load_data_from_file(file_path, key): with open(file_path) as f: return torch.FloatTensor(json.load(f)[key]) - train_transforms = Compose( - [ - monai.transforms.LoadImaged(keys=["image"]), - monai.transforms.EnsureChannelFirstd(keys=["image"]), + train_transforms_list = [ + monai.transforms.LoadImaged(keys=["image"]), + monai.transforms.EnsureChannelFirstd(keys=["image"]), + monai.transforms.Lambdad(keys="spacing", func=lambda x: _load_data_from_file(x, "spacing")), + monai.transforms.Lambdad(keys="spacing", func=lambda x: x * 1e2), + ] + if include_body_region: + train_transforms_list += [ monai.transforms.Lambdad( keys="top_region_index", func=lambda x: _load_data_from_file(x, "top_region_index") ), monai.transforms.Lambdad( keys="bottom_region_index", func=lambda x: _load_data_from_file(x, "bottom_region_index") ), - monai.transforms.Lambdad(keys="spacing", func=lambda x: _load_data_from_file(x, "spacing")), monai.transforms.Lambdad(keys="top_region_index", func=lambda x: x * 1e2), monai.transforms.Lambdad(keys="bottom_region_index", func=lambda x: x * 1e2), - monai.transforms.Lambdad(keys="spacing", func=lambda x: x * 1e2), ] - ) + train_transforms = Compose(train_transforms_list) train_ds = monai.data.CacheDataset( data=train_files, transform=train_transforms, cache_rate=cache_rate, num_workers=num_workers @@ -216,6 +225,9 @@ def train_one_epoch( Returns: torch.Tensor: Training loss for the epoch. """ + include_body_region = unet.include_top_region_index_input + include_modality = unet.num_class_embeds is not None + if local_rank == 0: current_lr = optimizer.param_groups[0]["lr"] logger.info(f"Epoch {epoch + 1}, lr {current_lr}.") @@ -231,30 +243,64 @@ def train_one_epoch( images = train_data["image"].to(device) images = images * scale_factor - top_region_index_tensor = train_data["top_region_index"].to(device) - bottom_region_index_tensor = train_data["bottom_region_index"].to(device) + if include_body_region: + top_region_index_tensor = train_data["top_region_index"].to(device) + bottom_region_index_tensor = train_data["bottom_region_index"].to(device) + # We trained with only CT in this version + if include_modality: + modality_tensor = torch.ones((len(images),), dtype=torch.long).to(device) spacing_tensor = train_data["spacing"].to(device) optimizer.zero_grad(set_to_none=True) with autocast("cuda", enabled=amp): - noise = torch.randn( - (num_images_per_batch, 4, images.size(-3), images.size(-2), images.size(-1)), device=device - ) + noise = torch.randn_like(images) - timesteps = torch.randint(0, num_train_timesteps, (images.shape[0],), device=images.device).long() + if isinstance(noise_scheduler, RFlowScheduler): + timesteps = noise_scheduler.sample_timesteps(images) + else: + timesteps = torch.randint(0, num_train_timesteps, (images.shape[0],), device=images.device).long() noisy_latent = noise_scheduler.add_noise(original_samples=images, noise=noise, timesteps=timesteps) - noise_pred = unet( - x=noisy_latent, - timesteps=timesteps, - top_region_index_tensor=top_region_index_tensor, - bottom_region_index_tensor=bottom_region_index_tensor, - spacing_tensor=spacing_tensor, - ) + # Create a dictionary to store the inputs + unet_inputs = { + "x": noisy_latent, + "timesteps": timesteps, + "spacing_tensor": spacing_tensor, + } + # Add extra arguments if include_body_region is True + if include_body_region: + unet_inputs.update( + { + "top_region_index_tensor": top_region_index_tensor, + "bottom_region_index_tensor": bottom_region_index_tensor, + } + ) + if include_modality: + unet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + model_output = unet(**unet_inputs) + + if noise_scheduler.prediction_type == DDPMPredictionType.EPSILON: + # predict noise + model_gt = noise + elif noise_scheduler.prediction_type == DDPMPredictionType.SAMPLE: + # predict sample + model_gt = images + elif noise_scheduler.prediction_type == DDPMPredictionType.V_PREDICTION: + # predict velocity + model_gt = images - noise + else: + raise ValueError( + "noise scheduler prediction type has to be chosen from ", + f"[{DDPMPredictionType.EPSILON},{DDPMPredictionType.SAMPLE},{DDPMPredictionType.V_PREDICTION}]", + ) - loss = loss_pt(noise_pred.float(), noise.float()) + loss = loss_pt(model_output.float(), model_gt.float()) if amp: scaler.scale(loss).backward() @@ -345,6 +391,10 @@ def diff_model_train( Path(args.model_dir).mkdir(parents=True, exist_ok=True) + unet = load_unet(args, device, logger) + noise_scheduler = define_instance(args, "noise_scheduler") + include_body_region = unet.include_top_region_index_input + filenames_train = load_filenames(args.json_data_list) if local_rank == 0: logger.info(f"num_files_train: {len(filenames_train)}") @@ -356,21 +406,24 @@ def diff_model_train( continue str_info = os.path.join(args.embedding_base_dir, filenames_train[_i]) + ".json" - train_files.append( - {"image": str_img, "top_region_index": str_info, "bottom_region_index": str_info, "spacing": str_info} - ) + train_files_i = {"image": str_img, "spacing": str_info} + if include_body_region: + train_files_i["top_region_index"] = str_info + train_files_i["bottom_region_index"] = str_info + train_files.append(train_files_i) if dist.is_initialized(): train_files = partition_dataset( data=train_files, shuffle=True, num_partitions=dist.get_world_size(), even_divisible=True )[local_rank] train_loader = prepare_data( - train_files, device, args.diffusion_unet_train["cache_rate"], batch_size=args.diffusion_unet_train["batch_size"] + train_files, + device, + args.diffusion_unet_train["cache_rate"], + batch_size=args.diffusion_unet_train["batch_size"], + include_body_region=include_body_region, ) - unet = load_unet(args, device, logger) - noise_scheduler = define_instance(args, "noise_scheduler") - scale_factor = calculate_scale_factor(train_loader, device, logger) optimizer = create_optimizer(unet, args.diffusion_unet_train["lr"]) diff --git a/generation/maisi/scripts/find_masks.py b/generation/maisi/scripts/find_masks.py index c919d3932f..b7a730c463 100644 --- a/generation/maisi/scripts/find_masks.py +++ b/generation/maisi/scripts/find_masks.py @@ -107,19 +107,21 @@ def find_masks( if not set(anatomy_list).issubset(_item["label_list"]): continue - # extract region indice (top_index and bottom_index) for candidate mask - top_index = [index for index, element in enumerate(_item["top_region_index"]) if element != 0] - top_index = top_index[0] - bottom_index = [index for index, element in enumerate(_item["bottom_region_index"]) if element != 0] - bottom_index = bottom_index[0] - # whether to keep this mask, default to be True. keep_mask = True - # if candiate mask does not contain all the body_region, skip it - for _idx in body_region: - if _idx > bottom_index or _idx < top_index: - keep_mask = False + # extract region indice (top_index and bottom_index) for candidate mask + include_body_region = "top_region_index" in _item.keys() + if include_body_region: + top_index = [index for index, element in enumerate(_item["top_region_index"]) if element != 0] + top_index = top_index[0] + bottom_index = [index for index, element in enumerate(_item["bottom_region_index"]) if element != 0] + bottom_index = bottom_index[0] + + # if candiate mask does not contain all the body_region, skip it + for _idx in body_region: + if _idx > bottom_index or _idx < top_index: + keep_mask = False for tumor_label in [23, 24, 26, 27, 128]: # we skip those mask with tumors if users do not provide tumor label in anatomy_list @@ -138,9 +140,10 @@ def find_masks( "pseudo_label": os.path.join(mask_foldername, _item["pseudo_label_filename"]), "spacing": _item["spacing"], "dim": _item["dim"], - "top_region_index": _item["top_region_index"], - "bottom_region_index": _item["bottom_region_index"], } + if include_body_region: + candidate["top_region_index"] = _item["top_region_index"] + candidate["bottom_region_index"] = _item["bottom_region_index"] # Conditionally add the label to the candidate dictionary if "label_filename" in _item: diff --git a/generation/maisi/scripts/infer_controlnet.py b/generation/maisi/scripts/infer_controlnet.py index 04ac982c98..0e88547d52 100644 --- a/generation/maisi/scripts/infer_controlnet.py +++ b/generation/maisi/scripts/infer_controlnet.py @@ -18,7 +18,7 @@ import torch import torch.distributed as dist -from monai.data import decollate_batch, MetaTensor +from monai.data import MetaTensor, decollate_batch from monai.networks.utils import copy_model_state from monai.transforms import SaveImage from monai.utils import RankFilter @@ -49,6 +49,7 @@ def main(): help="config json file that stores training hyper-parameters", ) parser.add_argument("-g", "--gpus", default=1, type=int, help="number of gpus per node") + args = parser.parse_args() # Step 0: configuration @@ -109,6 +110,9 @@ def main(): # define diffusion Model unet = define_instance(args, "diffusion_unet_def").to(device) + include_body_region = unet.include_top_region_index_input + include_modality = unet.num_class_embeds is not None + # load trained diffusion model if args.trained_diffusion_path is not None: if not os.path.exists(args.trained_diffusion_path): @@ -150,9 +154,14 @@ def main(): # get label mask labels = batch["label"].to(device) # get corresponding conditions - top_region_index_tensor = batch["top_region_index"].to(device) - bottom_region_index_tensor = batch["bottom_region_index"].to(device) + if include_body_region: + top_region_index_tensor = batch["top_region_index"].to(device) + bottom_region_index_tensor = batch["bottom_region_index"].to(device) + else: + top_region_index_tensor = None + bottom_region_index_tensor = None spacing_tensor = batch["spacing"].to(device) + modality_tensor = args.controlnet_infer["modality"] * torch.ones((len(labels),), dtype=torch.long).to(device) out_spacing = tuple((batch["spacing"].squeeze().numpy() / 100).tolist()) # get target dimension dim = batch["dim"] @@ -162,22 +171,23 @@ def main(): check_input(None, None, None, output_size, out_spacing, None) # generate a single synthetic image using a latent diffusion model with controlnet. synthetic_images, _ = ldm_conditional_sample_one_image( - autoencoder, - unet, - controlnet, - noise_scheduler, - scale_factor, - device, - labels, - top_region_index_tensor, - bottom_region_index_tensor, - spacing_tensor, + autoencoder=autoencoder, + diffusion_unet=unet, + controlnet=controlnet, + noise_scheduler=noise_scheduler, + scale_factor=scale_factor, + device=device, + combine_label_or=labels, + top_region_index_tensor=top_region_index_tensor, + bottom_region_index_tensor=bottom_region_index_tensor, + spacing_tensor=spacing_tensor, + modality_tensor=modality_tensor, latent_shape=latent_shape, output_size=output_size, noise_factor=1.0, num_inference_steps=args.controlnet_infer["num_inference_steps"], - # reduce it when GPU memory is limited autoencoder_sliding_window_infer_size=args.controlnet_infer["autoencoder_sliding_window_infer_size"], + autoencoder_sliding_window_infer_overlap=args.controlnet_infer["autoencoder_sliding_window_infer_overlap"], ) # save image/label pairs labels = decollate_batch(batch)[0]["label"] diff --git a/generation/maisi/scripts/inference.py b/generation/maisi/scripts/inference.py index 9049aee89a..3f81f9c49b 100644 --- a/generation/maisi/scripts/inference.py +++ b/generation/maisi/scripts/inference.py @@ -14,8 +14,8 @@ import json import logging import os -import tempfile import sys +import tempfile import monai import torch @@ -23,6 +23,7 @@ from monai.config import print_config from monai.transforms import LoadImage, Orientation from monai.utils import set_determinism + from scripts.sample import LDMSampler, check_input from scripts.utils import define_instance from scripts.utils_plot import find_label_center_loc, get_xyz_plot, show_image @@ -60,10 +61,18 @@ def main(): default=None, help="random seed, can be None or int", ) + parser.add_argument( + "--version", + default="maisi3d-rflow", + type=str, + help="maisi_version, choose from ['maisi3d-ddpm', 'maisi3d-rflow']", + ) args = parser.parse_args() # Step 0: configuration logger = logging.getLogger("maisi.inference") + maisi_version = args.version + # ## Set deterministic training for reproducibility if args.random_seed is not None: set_determinism(seed=args.random_seed) @@ -79,41 +88,75 @@ def main(): root_dir = tempfile.mkdtemp() if directory is None else directory print(root_dir) + # TODO: remove the `files` after the files are uploaded to the NGC files = [ { "path": "models/autoencoder_epoch273.pt", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo/model_maisi_autoencoder_epoch273_alternative.pt", - }, - { - "path": "models/input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1.pt", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo/model_maisi_input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1_alternative.pt", - }, - { - "path": "models/controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current.pt", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo/model_maisi_controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current_alternative.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials" + "/model_zoo/model_maisi_autoencoder_epoch273_alternative.pt", }, { "path": "models/mask_generation_autoencoder.pt", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/mask_generation_autoencoder.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai" + "/tutorials/mask_generation_autoencoder.pt", }, { "path": "models/mask_generation_diffusion_unet.pt", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo/model_maisi_mask_generation_diffusion_unet_v2.pt", - }, - { - "path": "configs/candidate_masks_flexible_size_and_spacing_3000.json", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/candidate_masks_flexible_size_and_spacing_3000.json", + "url": "https://developer.download.nvidia.com/assets/Clara/monai" + "/tutorials/model_zoo/model_maisi_mask_generation_diffusion_unet_v2.pt", }, { "path": "configs/all_anatomy_size_condtions.json", "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/all_anatomy_size_condtions.json", }, { - "path": "datasets/all_masks_flexible_size_and_spacing_3000.zip", - "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo/model_maisi_all_masks_flexible_size_and_spacing_3000.zip", + "path": "datasets/all_masks_flexible_size_and_spacing_4000.zip", + "url": "https://developer.download.nvidia.com/assets/Clara/monai" + "/tutorials/all_masks_flexible_size_and_spacing_4000.zip", }, ] + if maisi_version == "maisi3d-ddpm": + files += [ + { + "path": "models/diff_unet_3d_ddpm.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo" + "/model_maisi_input_unet3d_data-all_steps1000size512ddpm_random_current_inputx_v1_alternative.pt", + }, + { + "path": "models/controlnet_3d_ddpm.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/model_zoo" + "/model_maisi_controlnet-20datasets-e20wl100fold0bc_noi_dia_fsize_current_alternative.pt", + }, + { + "path": "configs/candidate_masks_flexible_size_and_spacing_3000.json", + "url": "https://developer.download.nvidia.com/assets/Clara/monai" + "/tutorials/candidate_masks_flexible_size_and_spacing_3000.json", + }, + ] + elif maisi_version == "maisi3d-rflow": + files += [ + { + "path": "models/diff_unet_3d_rflow.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/" + "diff_unet_ckpt_rflow_epoch19350.pt", + }, + { + "path": "models/controlnet_3d_rflow.pt", + "url": "https://developer.download.nvidia.com/assets/Clara/monai/tutorials/" + "controlnet_rflow_epoch60.pt", + }, + { + "path": "configs/candidate_masks_flexible_size_and_spacing_4000.json", + "url": "https://developer.download.nvidia.com/assets/Clara/monai" + "/tutorials/candidate_masks_flexible_size_and_spacing_4000.json", + }, + ] + else: + raise ValueError( + f"maisi_version has to be chosen from ['maisi3d-ddpm', 'maisi3d-rflow'], yet got {maisi_version}." + ) + for file in files: file["path"] = file["path"] if "datasets/" not in file["path"] else os.path.join(root_dir, file["path"]) download_url(url=file["url"], filepath=file["path"]) @@ -230,6 +273,7 @@ def main(): image_output_ext=args.image_output_ext, label_output_ext=args.label_output_ext, spacing=args.spacing, + modality=args.modality, num_inference_steps=args.num_inference_steps, mask_generation_num_inference_steps=args.mask_generation_num_inference_steps, random_seed=args.random_seed, diff --git a/generation/maisi/scripts/quality_check.py b/generation/maisi/scripts/quality_check.py index 223732761a..bff49b6da0 100644 --- a/generation/maisi/scripts/quality_check.py +++ b/generation/maisi/scripts/quality_check.py @@ -9,7 +9,6 @@ # See the License for the specific language governing permissions and # limitations under the License. -import nibabel as nib import numpy as np @@ -110,8 +109,11 @@ def is_outlier(statistics, image_data, label_data, label_int_dict): for label_name, stats in statistics.items(): # Get the thresholds from the statistics - low_thresh = stats["sigma_6_low"] # or "sigma_12_low" depending on your needs - high_thresh = stats["sigma_6_high"] # or "sigma_12_high" depending on your needs + low_thresh = min(stats["sigma_6_low"], stats["percentile_0_5"]) # or "sigma_12_low" depending on your needs + high_thresh = max(stats["sigma_6_high"], stats["percentile_99_5"]) # or "sigma_12_high" depending on your needs + + if label_name == "bone": + high_thresh = 1000.0 # Retrieve the corresponding label integers labels = label_int_dict.get(label_name, []) diff --git a/generation/maisi/scripts/sample.py b/generation/maisi/scripts/sample.py index c1e2c8699a..fb1d0f4251 100644 --- a/generation/maisi/scripts/sample.py +++ b/generation/maisi/scripts/sample.py @@ -16,20 +16,29 @@ import random import time from datetime import datetime +import warnings +import gc import monai import torch -from monai.inferers.inferer import DiffusionInferer from monai.data import MetaTensor -from monai.inferers import sliding_window_inference +from monai.inferers.inferer import DiffusionInferer from monai.transforms import Compose, SaveImage from monai.utils import set_determinism from tqdm import tqdm +from monai.inferers.inferer import SlidingWindowInferer +from monai.networks.schedulers import RFlowScheduler, DDPMScheduler from .augmentation import augmentation from .find_masks import find_masks -from .utils import binarize_labels, general_mask_generation_post_process, get_body_region_index_from_mask, remap_labels from .quality_check import is_outlier +from .utils import ( + binarize_labels, + general_mask_generation_post_process, + get_body_region_index_from_mask, + remap_labels, + dynamic_infer, +) class ReconModel(torch.nn.Module): @@ -122,7 +131,19 @@ def ldm_conditional_sample_one_mask( latents = initialize_noise_latents(latent_shape, device) anatomy_size = torch.FloatTensor(anatomy_size).unsqueeze(0).unsqueeze(0).half().to(device) # synthesize latents + if isinstance(noise_scheduler, DDPMScheduler) and num_inference_steps < noise_scheduler.num_train_timesteps: + warnings.warn( + "**************************************************************\n" + "* WARNING: Mask noise_scheduler is a DDPMScheduler.\n" + "* We expect num_inference_steps = noise_scheduler.num_train_timesteps" + f" = {noise_scheduler.num_train_timesteps}.\n" + f"* Yet got num_inference_steps = {num_inference_steps}.\n" + "* The generated image quality is not guaranteed.\n" + "**************************************************************" + ) + noise_scheduler.set_timesteps(num_inference_steps=num_inference_steps) + # mask generator is DDPM inferer_ddpm = DiffusionInferer(noise_scheduler) latents = inferer_ddpm.sample( input_noise=latents, @@ -131,29 +152,17 @@ def ldm_conditional_sample_one_mask( verbose=True, conditioning=anatomy_size.to(device), ) - # decode latents to synthesized masks - if math.prod(latent_shape[1:]) <= math.prod(autoencoder_sliding_window_infer_size): - synthetic_mask = recon_model(latents).cpu().detach() - else: - synthetic_mask = ( - sliding_window_inference( - inputs=latents, - roi_size=( - autoencoder_sliding_window_infer_size[0], - autoencoder_sliding_window_infer_size[1], - autoencoder_sliding_window_infer_size[2], - ), - sw_batch_size=1, - predictor=recon_model, - mode="gaussian", - overlap=autoencoder_sliding_window_infer_overlap, - sw_device=device, - device=torch.device("cpu"), - progress=True, - ) - .cpu() - .detach() - ) + + inferer = SlidingWindowInferer( + roi_size=autoencoder_sliding_window_infer_size, + sw_batch_size=1, + progress=True, + mode="gaussian", + overlap=autoencoder_sliding_window_infer_overlap, + sw_device=device, + device=torch.device("cpu"), + ) + synthetic_mask = dynamic_infer(inferer, recon_model, latents) synthetic_mask = torch.softmax(synthetic_mask, dim=1) synthetic_mask = torch.argmax(synthetic_mask, dim=1, keepdim=True) # mapping raw index to 132 labels @@ -183,12 +192,13 @@ def ldm_conditional_sample_one_image( scale_factor, device, combine_label_or, - top_region_index_tensor, - bottom_region_index_tensor, spacing_tensor, latent_shape, output_size, noise_factor, + top_region_index_tensor=None, + bottom_region_index_tensor=None, + modality_tensor=None, num_inference_steps=1000, autoencoder_sliding_window_infer_size=[96, 96, 96], autoencoder_sliding_window_infer_overlap=0.6667, @@ -204,12 +214,13 @@ def ldm_conditional_sample_one_image( scale_factor (float): Scaling factor for the latent space. device (torch.device): The device to run the computation on. combine_label_or (torch.Tensor): The combined label tensor. - top_region_index_tensor (torch.Tensor): Tensor specifying the top region index. - bottom_region_index_tensor (torch.Tensor): Tensor specifying the bottom region index. spacing_tensor (torch.Tensor): Tensor specifying the spacing. latent_shape (tuple): The shape of the latent space. output_size (tuple): The desired output size of the image. noise_factor (float): Factor to scale the initial noise. + top_region_index_tensor (torch.Tensor): Tensor specifying the top region index. Defaults to None. + bottom_region_index_tensor (torch.Tensor): Tensor specifying the bottom region index. Defaults to None. + modality_tensor (torch.Tensor): Int Tensor specifying the modality. num_inference_steps (int): Number of inference steps for the diffusion process. autoencoder_sliding_window_infer_size (list, optional): Size of the sliding window for inference. Defaults to [96, 96, 96]. autoencoder_sliding_window_infer_overlap (float, optional): Overlap ratio for sliding window inference. Defaults to 0.6667. @@ -224,6 +235,9 @@ def ldm_conditional_sample_one_image( b_min = 0.0 b_max = 1 + include_body_region = diffusion_unet.include_top_region_index_input + include_modality = diffusion_unet.num_class_embeds is not None + recon_model = ReconModel(autoencoder=autoencoder, scale_factor=scale_factor).to(device) with torch.no_grad(), torch.amp.autocast("cuda"): @@ -247,54 +261,106 @@ def ldm_conditional_sample_one_image( latents = initialize_noise_latents(latent_shape, device) * noise_factor # synthesize latents - noise_scheduler.set_timesteps(num_inference_steps=num_inference_steps) - for t in tqdm(noise_scheduler.timesteps, ncols=110): - # Get controlnet output - down_block_res_samples, mid_block_res_sample = controlnet( - x=latents, - timesteps=torch.Tensor((t,)).to(device), - controlnet_cond=controlnet_cond_vis, + if isinstance(noise_scheduler, RFlowScheduler): + noise_scheduler.set_timesteps( + num_inference_steps=num_inference_steps, + input_img_size_numel=torch.prod(torch.tensor(latents.shape[2:])), ) - latent_model_input = latents - noise_pred = diffusion_unet( - x=latent_model_input, - timesteps=torch.Tensor((t,)).to(device), - top_region_index_tensor=top_region_index_tensor, - bottom_region_index_tensor=bottom_region_index_tensor, - spacing_tensor=spacing_tensor, - down_block_additional_residuals=down_block_res_samples, - mid_block_additional_residual=mid_block_res_sample, + else: + noise_scheduler.set_timesteps(num_inference_steps=num_inference_steps) + + if isinstance(noise_scheduler, DDPMScheduler) and num_inference_steps < noise_scheduler.num_train_timesteps: + warnings.warn( + "**************************************************************\n" + "* WARNING: Image noise_scheduler is a DDPMScheduler.\n" + "* We expect num_inference_steps = noise_scheduler.num_train_timesteps" + f" = {noise_scheduler.num_train_timesteps}.\n" + f"* Yet got num_inference_steps = {num_inference_steps}.\n" + "* The generated image quality is not guaranteed.\n" + "**************************************************************" ) - latents, _ = noise_scheduler.step(noise_pred, t, latents) + + all_timesteps = noise_scheduler.timesteps + all_next_timesteps = torch.cat((all_timesteps[1:], torch.tensor([0], dtype=all_timesteps.dtype))) + progress_bar = tqdm( + zip(all_timesteps, all_next_timesteps), + total=min(len(all_timesteps), len(all_next_timesteps)), + ) + for t, next_t in progress_bar: + # get controlnet output + # Create a dictionary to store the inputs + controlnet_inputs = { + "x": latents, + "timesteps": torch.Tensor((t,)).to(device), + "controlnet_cond": controlnet_cond_vis, + } + if include_modality: + controlnet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + down_block_res_samples, mid_block_res_sample = controlnet(**controlnet_inputs) + + # get diffusion network output + # Create a dictionary to store the inputs + unet_inputs = { + "x": latents, + "timesteps": torch.Tensor((t,)).to(device), + "spacing_tensor": spacing_tensor, + "down_block_additional_residuals": down_block_res_samples, + "mid_block_additional_residual": mid_block_res_sample, + } + # Add extra arguments if include_body_region is True + if include_body_region: + unet_inputs.update( + { + "top_region_index_tensor": top_region_index_tensor, + "bottom_region_index_tensor": bottom_region_index_tensor, + } + ) + if include_modality: + unet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + model_output = diffusion_unet(**unet_inputs) + + if not isinstance(noise_scheduler, RFlowScheduler): + latents, _ = noise_scheduler.step(model_output, t, latents) # type: ignore + else: + latents, _ = noise_scheduler.step(model_output, t, latents, next_t) # type: ignore end_time = time.time() - logging.info(f"---- Latent features generation time: {end_time - start_time} seconds ----") - del noise_pred + logging.info(f"---- DM/ControlNet Latent features generation time: {end_time - start_time} seconds ----") + del ( + unet_inputs, + controlnet_inputs, + model_output, + controlnet_cond_vis, + down_block_res_samples, + mid_block_res_sample, + ) + gc.collect() torch.cuda.empty_cache() # decode latents to synthesized images logging.info("---- Start decoding latent features into images... ----") start_time = time.time() - if math.prod(latent_shape[1:]) <= math.prod(autoencoder_sliding_window_infer_size): - synthetic_images = recon_model(latents) - else: - synthetic_images = sliding_window_inference( - inputs=latents, - roi_size=( - min(output_size[0] // 4, autoencoder_sliding_window_infer_size[0]), - min(output_size[1] // 4, autoencoder_sliding_window_infer_size[1]), - min(output_size[2] // 4, autoencoder_sliding_window_infer_size[2]), - ), - sw_batch_size=1, - predictor=recon_model, - mode="gaussian", - overlap=autoencoder_sliding_window_infer_overlap, - sw_device=device, - device=torch.device("cpu"), - progress=True, - ) + + inferer = SlidingWindowInferer( + roi_size=autoencoder_sliding_window_infer_size, + sw_batch_size=1, + progress=True, + mode="gaussian", + overlap=autoencoder_sliding_window_infer_overlap, + sw_device=device, + device=torch.device("cpu"), + ) + synthetic_images = dynamic_infer(inferer, recon_model, latents) synthetic_images = torch.clip(synthetic_images, b_min, b_max).cpu() end_time = time.time() - logging.info(f"---- Image decoding time: {end_time - start_time} seconds ----") + logging.info(f"---- Image VAE decoding time: {end_time - start_time} seconds ----") ## post processing: # project output to [0, 1] @@ -510,6 +576,7 @@ def __init__( label_output_ext=".nii.gz", real_img_median_statistics="./configs/image_median_statistics.json", spacing=[1, 1, 1], + modality=1, num_inference_steps=None, mask_generation_num_inference_steps=None, random_seed=None, @@ -522,6 +589,7 @@ def __init__( Args: Various parameters related to model configuration, input settings, and output specifications. """ + self.random_seed = random_seed if random_seed is not None: set_determinism(seed=random_seed) @@ -575,7 +643,7 @@ def __init__( self.autoencoder_sliding_window_infer_overlap = autoencoder_sliding_window_infer_overlap # quality check args - self.max_try_time = 5 # if not pass quality check, will try self.max_try_time times + self.max_try_time = 2 # if not pass quality check, will try self.max_try_time times with open(real_img_median_statistics, "r") as json_file: self.median_statistics = json.load(json_file) self.label_int_dict = { @@ -601,21 +669,27 @@ def __init__( self.mask_generation_diffusion_unet.eval() self.spacing = spacing - - self.val_transforms = Compose( - [ - monai.transforms.LoadImaged(keys=["pseudo_label"]), - monai.transforms.EnsureChannelFirstd(keys=["pseudo_label"]), - monai.transforms.Orientationd(keys=["pseudo_label"], axcodes="RAS"), - monai.transforms.EnsureTyped(keys=["pseudo_label"], dtype=torch.uint8), + self.modality_tensor = modality * torch.ones((1,), dtype=torch.long).to(device) + self.include_body_region = self.diffusion_unet.include_top_region_index_input + self.include_modality = self.diffusion_unet.num_class_embeds is not None + + val_transforms_list = [ + monai.transforms.LoadImaged(keys=["pseudo_label"]), + monai.transforms.EnsureChannelFirstd(keys=["pseudo_label"]), + monai.transforms.Orientationd(keys=["pseudo_label"], axcodes="RAS"), + monai.transforms.EnsureTyped(keys=["pseudo_label"], dtype=torch.uint8), + monai.transforms.Lambdad(keys="spacing", func=lambda x: torch.FloatTensor(x)), + monai.transforms.Lambdad(keys="spacing", func=lambda x: x * 1e2), + ] + if self.include_body_region: + val_transforms_list += [ monai.transforms.Lambdad(keys="top_region_index", func=lambda x: torch.FloatTensor(x)), monai.transforms.Lambdad(keys="bottom_region_index", func=lambda x: torch.FloatTensor(x)), - monai.transforms.Lambdad(keys="spacing", func=lambda x: torch.FloatTensor(x)), monai.transforms.Lambdad(keys="top_region_index", func=lambda x: x * 1e2), monai.transforms.Lambdad(keys="bottom_region_index", func=lambda x: x * 1e2), - monai.transforms.Lambdad(keys="spacing", func=lambda x: x * 1e2), ] - ) + + self.val_transforms = Compose(val_transforms_list) logging.info("LDM sampler initialized.") def sample_multiple_images(self, num_img): @@ -625,6 +699,7 @@ def sample_multiple_images(self, num_img): Args: num_img (int): Number of images to generate. """ + modality_tensor = self.modality_tensor output_filenames = [] if len(self.controllable_anatomy_size) > 0: # we will use mask generation instead of finding candidate masks @@ -653,11 +728,19 @@ def sample_multiple_images(self, num_img): selected_mask_files = self.select_mask(candidate_mask_files, num_img) logging.info(f"Images will be generated based on {selected_mask_files}.") - if len(selected_mask_files) != num_img: + if len(selected_mask_files) < num_img: raise ValueError( - f"len(selected_mask_files) ({len(selected_mask_files)}) != num_img ({num_img}). This should not happen. Please revisit function select_mask(self, candidate_mask_files, num_img)." + ( + f"len(selected_mask_files) ({len(selected_mask_files)}) < num_img ({num_img}). " + "This should not happen. Please revisit function select_mask(self, candidate_mask_files, num_img)." + ) ) - for item in selected_mask_files: + + num_generated_img = 0 + for index_s in range(len(selected_mask_files)): + item = selected_mask_files[index_s] + if num_generated_img >= num_img: + break logging.info("---- Start preparing masks... ----") start_time = time.time() if len(self.controllable_anatomy_size) > 0: @@ -682,58 +765,63 @@ def sample_multiple_images(self, num_img): combine_label_or = self.ensure_output_size_and_spacing(combine_label_or) # mask augmentation if if_aug: - combine_label_or = augmentation(combine_label_or, self.output_size) + combine_label_or = augmentation(combine_label_or, self.output_size, self.random_seed) end_time = time.time() logging.info(f"---- Mask preparation time: {end_time - start_time} seconds ----") torch.cuda.empty_cache() # generate image/label pairs to_generate = True try_time = 0 - while to_generate: - synthetic_images, synthetic_labels = self.sample_one_pair( - combine_label_or, - top_region_index_tensor, - bottom_region_index_tensor, - spacing_tensor, - ) - # synthetic image quality check - pass_quality_check = self.quality_check( - synthetic_images.cpu().detach().numpy(), combine_label_or.cpu().detach().numpy() - ) - if pass_quality_check or try_time > self.max_try_time: - # save image/label pairs - output_postfix = datetime.now().strftime("%Y%m%d_%H%M%S_%f") - synthetic_labels.meta["filename_or_obj"] = "sample.nii.gz" - synthetic_images = MetaTensor(synthetic_images, meta=synthetic_labels.meta) - img_saver = SaveImage( - output_dir=self.output_dir, - output_postfix=output_postfix + "_image", - output_ext=self.image_output_ext, - separate_folder=False, - ) - img_saver(synthetic_images[0]) - synthetic_images_filename = os.path.join( - self.output_dir, "sample_" + output_postfix + "_image" + self.image_output_ext - ) - # filter out the organs that are not in anatomy_list - synthetic_labels = filter_mask_with_organs(synthetic_labels, self.anatomy_list) - label_saver = SaveImage( - output_dir=self.output_dir, - output_postfix=output_postfix + "_label", - output_ext=self.label_output_ext, - separate_folder=False, - ) - label_saver(synthetic_labels[0]) - synthetic_labels_filename = os.path.join( - self.output_dir, "sample_" + output_postfix + "_label" + self.label_output_ext - ) - output_filenames.append([synthetic_images_filename, synthetic_labels_filename]) - to_generate = False - else: + # start generation + synthetic_images, synthetic_labels = self.sample_one_pair( + combine_label_or, + top_region_index_tensor, + bottom_region_index_tensor, + spacing_tensor, + modality_tensor, + ) + # synthetic image quality check + pass_quality_check = self.quality_check( + synthetic_images.cpu().detach().numpy(), combine_label_or.cpu().detach().numpy() + ) + print(num_img - num_generated_img, (len(selected_mask_files) - index_s)) + if pass_quality_check or (num_img - num_generated_img) >= (len(selected_mask_files) - index_s): + if not pass_quality_check: logging.info( - "Generated image/label pair did not pass quality check, will re-generate another pair." + "Generated image/label pair did not pass quality check, but will still save them. " + "Please consider changing spacing and output_size to facilitate a more realistic setting." ) - try_time += 1 + num_generated_img = num_generated_img + 1 + # save image/label pairs + output_postfix = datetime.now().strftime("%Y%m%d_%H%M%S_%f") + synthetic_labels.meta["filename_or_obj"] = "sample.nii.gz" + synthetic_images = MetaTensor(synthetic_images, meta=synthetic_labels.meta) + img_saver = SaveImage( + output_dir=self.output_dir, + output_postfix=output_postfix + "_image", + output_ext=self.image_output_ext, + separate_folder=False, + ) + img_saver(synthetic_images[0]) + synthetic_images_filename = os.path.join( + self.output_dir, "sample_" + output_postfix + "_image" + self.image_output_ext + ) + # filter out the organs that are not in anatomy_list + synthetic_labels = filter_mask_with_organs(synthetic_labels, self.anatomy_list) + label_saver = SaveImage( + output_dir=self.output_dir, + output_postfix=output_postfix + "_label", + output_ext=self.label_output_ext, + separate_folder=False, + ) + label_saver(synthetic_labels[0]) + synthetic_labels_filename = os.path.join( + self.output_dir, "sample_" + output_postfix + "_label" + self.label_output_ext + ) + output_filenames.append([synthetic_images_filename, synthetic_labels_filename]) + to_generate = False + else: + logging.info("Generated image/label pair did not pass quality check, will re-generate another pair.") return output_filenames def select_mask(self, candidate_mask_files, num_img): @@ -750,7 +838,7 @@ def select_mask(self, candidate_mask_files, num_img): selected_mask_files = [] random.shuffle(candidate_mask_files) - for n in range(num_img): + for n in range(len(candidate_mask_files)): mask_file = candidate_mask_files[n % len(candidate_mask_files)] selected_mask_files.append({"mask_file": mask_file, "if_aug": True}) return selected_mask_files @@ -761,6 +849,7 @@ def sample_one_pair( top_region_index_tensor, bottom_region_index_tensor, spacing_tensor, + modality_tensor, ): """ Generate a single pair of synthetic image and mask. @@ -770,6 +859,7 @@ def sample_one_pair( top_region_index_tensor (torch.Tensor): Tensor specifying the top region index. bottom_region_index_tensor (torch.Tensor): Tensor specifying the bottom region index. spacing_tensor (torch.Tensor): Tensor specifying the spacing. + modality_tensor (torch.Tensor): Int Tensor specifying the modality. Returns: tuple: A tuple containing the synthetic image and its corresponding label. @@ -786,6 +876,7 @@ def sample_one_pair( top_region_index_tensor=top_region_index_tensor, bottom_region_index_tensor=bottom_region_index_tensor, spacing_tensor=spacing_tensor, + modality_tensor=modality_tensor, latent_shape=self.latent_shape, output_size=self.output_size, noise_factor=self.noise_factor, @@ -959,13 +1050,11 @@ def read_mask_information(self, mask_file): """ val_data = self.val_transforms(mask_file) - for key in [ - "pseudo_label", - "spacing", - "top_region_index", - "bottom_region_index", - ]: - val_data[key] = val_data[key].unsqueeze(0).to(self.device) + for key in ["pseudo_label", "spacing", "top_region_index", "bottom_region_index"]: + if isinstance(val_data[key], torch.Tensor): + val_data[key] = val_data[key].unsqueeze(0).to(self.device) + else: + val_data[key] = None return ( val_data["pseudo_label"], @@ -1000,42 +1089,75 @@ def find_closest_masks(self, num_img): if len(candidates) < num_img: raise ValueError(f"candidate masks are less than {num_img}).") + # loop through the database and find closest combinations new_candidates = [] for c in candidates: diff = 0 + include_c = True for axis in range(3): + if abs(c["dim"][axis]) < self.output_size[axis] - 64: + # we cannot upsample the mask too much + include_c = False + break + # check diff in FOV, major metric + diff += abs( + (abs(c["dim"][axis] * c["spacing"][axis]) - self.output_size[axis] * self.spacing[axis]) / 10 + ) # check diff in dim - diff += abs((c["dim"][axis] - self.output_size[axis]) / 100) + diff += abs((abs(c["dim"][axis]) - self.output_size[axis]) / 100) # check diff in spacing - diff += abs(c["spacing"][axis] - self.spacing[axis]) - new_candidates.append((c, diff)) + diff += abs(abs(c["spacing"][axis]) - self.spacing[axis]) + if include_c: + new_candidates.append((c, diff)) + # choose top-2*num_img candidates (at least 5) - new_candidates = sorted(new_candidates, key=lambda x: x[1])[: max(2 * num_img, 5)] + num_candidates = max(self.max_try_time * num_img, 5) + new_candidates = sorted(new_candidates, key=lambda x: x[1]) + final_candidates = [] # check top-2*num_img candidates and update spacing after resampling - image_loader = monai.transforms.LoadImage(image_only=True, ensure_channel_first=True) for c, _ in new_candidates: - label = image_loader(c["pseudo_label"]) - try: - label = self.ensure_output_size_and_spacing(label.unsqueeze(0)) - except ValueError as e: - if "Resampled mask does not contain required class labels" in str(e): - continue - else: - raise e - # get region_index after resample - top_region_index, bottom_region_index = get_body_region_index_from_mask(label) - c["top_region_index"] = top_region_index - c["bottom_region_index"] = bottom_region_index - c["spacing"] = self.spacing - c["dim"] = self.output_size - - final_candidates.append(c) + c = self.resample_mask_check_organ_list(c) + if c is not None: + final_candidates.append(c) + if len(final_candidates) >= num_candidates: + break if len(final_candidates) == 0: raise ValueError("Cannot find body region with given organ list.") return final_candidates + def resample_mask_check_organ_list(self, mask): + """ + Resample mask and check if the resampled mask contains the required organ list. + + Args: + mask (dict): input mask. + + Returns: + dict: resampled mask. If None, means the resampled mask does not contain the required organ list + + Raises: + ValueError: If suitable candidates cannot be found. + """ + + image_loader = monai.transforms.LoadImage(image_only=True, ensure_channel_first=True) + label = image_loader(mask["pseudo_label"]) + try: + label = self.ensure_output_size_and_spacing(label.unsqueeze(0)) + except ValueError as e: + if "Resampled mask does not contain required class labels" in str(e): + return None + else: + raise e + # get region_index after resample + top_region_index, bottom_region_index = get_body_region_index_from_mask(label) + mask["top_region_index"] = top_region_index + mask["bottom_region_index"] = bottom_region_index + mask["spacing"] = self.spacing + mask["dim"] = self.output_size + return mask + def quality_check(self, image_data, label_data): """ Perform a quality check on the generated image. diff --git a/generation/maisi/scripts/train_controlnet.py b/generation/maisi/scripts/train_controlnet.py index c59bebccff..3d7336be17 100644 --- a/generation/maisi/scripts/train_controlnet.py +++ b/generation/maisi/scripts/train_controlnet.py @@ -23,6 +23,8 @@ import torch.nn.functional as F from monai.networks.utils import copy_model_state from monai.utils import RankFilter +from monai.networks.schedulers import RFlowScheduler +from monai.networks.schedulers.ddpm import DDPMPredictionType from torch.amp import GradScaler, autocast from torch.nn.parallel import DistributedDataParallel as DDP from torch.utils.tensorboard import SummaryWriter @@ -51,6 +53,7 @@ def main(): help="config json file that stores training hyper-parameters", ) parser.add_argument("-g", "--gpus", default=1, type=int, help="number of gpus per node") + args = parser.parse_args() # Step 0: configuration @@ -105,6 +108,9 @@ def main(): # Step 2: define diffusion model and controlnet # define diffusion Model unet = define_instance(args, "diffusion_unet_def").to(device) + include_body_region = unet.include_top_region_index_input + include_modality = unet.num_class_embeds is not None + # load trained diffusion model if args.trained_diffusion_path is not None: if not os.path.exists(args.trained_diffusion_path): @@ -168,57 +174,104 @@ def main(): epoch_loss_ = 0 for step, batch in enumerate(train_loader): # get image embedding and label mask and scale image embedding by the provided scale_factor - inputs = batch["image"].to(device) * scale_factor + images = batch["image"].to(device) * scale_factor labels = batch["label"].to(device) # get corresponding conditions - top_region_index_tensor = batch["top_region_index"].to(device) - bottom_region_index_tensor = batch["bottom_region_index"].to(device) + if include_body_region: + top_region_index_tensor = batch["top_region_index"].to(device) + bottom_region_index_tensor = batch["bottom_region_index"].to(device) + # We trained with only CT in this version + if include_modality: + modality_tensor = torch.ones((len(images),), dtype=torch.long).to(device) spacing_tensor = batch["spacing"].to(device) optimizer.zero_grad(set_to_none=True) with autocast("cuda", enabled=True): # generate random noise - noise_shape = list(inputs.shape) - noise = torch.randn(noise_shape, dtype=inputs.dtype).to(device) + noise_shape = list(images.shape) + noise = torch.randn(noise_shape, dtype=images.dtype).to(device) # use binary encoding to encode segmentation mask controlnet_cond = binarize_labels(labels.as_tensor().to(torch.uint8)).float() # create timesteps - timesteps = torch.randint( - 0, noise_scheduler.num_train_timesteps, (inputs.shape[0],), device=device - ).long() + if isinstance(noise_scheduler, RFlowScheduler): + timesteps = noise_scheduler.sample_timesteps(images) + else: + timesteps = torch.randint( + 0, noise_scheduler.num_train_timesteps, (images.shape[0],), device=images.device + ).long() # create noisy latent - noisy_latent = noise_scheduler.add_noise(original_samples=inputs, noise=noise, timesteps=timesteps) + noisy_latent = noise_scheduler.add_noise(original_samples=images, noise=noise, timesteps=timesteps) # get controlnet output - down_block_res_samples, mid_block_res_sample = controlnet( - x=noisy_latent, timesteps=timesteps, controlnet_cond=controlnet_cond - ) - # get noise prediction from diffusion unet - noise_pred = unet( - x=noisy_latent, - timesteps=timesteps, - top_region_index_tensor=top_region_index_tensor, - bottom_region_index_tensor=bottom_region_index_tensor, - spacing_tensor=spacing_tensor, - down_block_additional_residuals=down_block_res_samples, - mid_block_additional_residual=mid_block_res_sample, + # Create a dictionary to store the inputs + controlnet_inputs = { + "x": noisy_latent, + "timesteps": timesteps, + "controlnet_cond": controlnet_cond, + } + if include_modality: + controlnet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + down_block_res_samples, mid_block_res_sample = controlnet(**controlnet_inputs) + + # get diffusion network output + # Create a dictionary to store the inputs + unet_inputs = { + "x": noisy_latent, + "timesteps": timesteps, + "spacing_tensor": spacing_tensor, + "down_block_additional_residuals": down_block_res_samples, + "mid_block_additional_residual": mid_block_res_sample, + } + # Add extra arguments if include_body_region is True + if include_body_region: + unet_inputs.update( + { + "top_region_index_tensor": top_region_index_tensor, + "bottom_region_index_tensor": bottom_region_index_tensor, + } + ) + if include_modality: + unet_inputs.update( + { + "class_labels": modality_tensor, + } + ) + model_output = unet(**unet_inputs) + + if noise_scheduler.prediction_type == DDPMPredictionType.EPSILON: + # predict noise + model_gt = noise + elif noise_scheduler.prediction_type == DDPMPredictionType.SAMPLE: + # predict sample + model_gt = images + elif noise_scheduler.prediction_type == DDPMPredictionType.V_PREDICTION: + # predict velocity + model_gt = images - noise + else: + raise ValueError( + "noise scheduler prediction type has to be chosen from ", + f"[{DDPMPredictionType.EPSILON},{DDPMPredictionType.SAMPLE},{DDPMPredictionType.V_PREDICTION}]", ) if weighted_loss > 1.0: - weights = torch.ones_like(inputs).to(inputs.device) - roi = torch.zeros([noise_shape[0]] + [1] + noise_shape[2:]).to(inputs.device) - interpolate_label = F.interpolate(labels, size=inputs.shape[2:], mode="nearest") + weights = torch.ones_like(images).to(images.device) + roi = torch.zeros([noise_shape[0]] + [1] + noise_shape[2:]).to(images.device) + interpolate_label = F.interpolate(labels, size=images.shape[2:], mode="nearest") # assign larger weights for ROI (tumor) for label in weighted_loss_label: roi[interpolate_label == label] = 1 - weights[roi.repeat(1, inputs.shape[1], 1, 1, 1) == 1] = weighted_loss - loss = (F.l1_loss(noise_pred.float(), noise.float(), reduction="none") * weights).mean() + weights[roi.repeat(1, images.shape[1], 1, 1, 1) == 1] = weighted_loss + loss = (F.l1_loss(noise_pred.float(), model_gt.float(), reduction="none") * weights).mean() else: - loss = F.l1_loss(noise_pred.float(), noise.float()) + loss = F.l1_loss(model_output.float(), model_gt.float()) scaler.scale(loss).backward() scaler.step(optimizer) diff --git a/generation/maisi/scripts/utils.py b/generation/maisi/scripts/utils.py index 13d9a240e1..9b1df28921 100644 --- a/generation/maisi/scripts/utils.py +++ b/generation/maisi/scripts/utils.py @@ -11,9 +11,9 @@ import copy import json +import logging import math import os -import logging from argparse import Namespace from datetime import timedelta from typing import Any, Sequence @@ -22,11 +22,11 @@ import skimage import torch import torch.distributed as dist -from monai.transforms.utils_morphological_ops import dilate, erode from monai.bundle import ConfigParser from monai.config import DtypeLike, NdarrayOrTensor from monai.data import CacheDataset, DataLoader, partition_dataset from monai.transforms import Compose, EnsureTyped, Lambdad, LoadImaged, Orientationd +from monai.transforms.utils_morphological_ops import dilate, erode from monai.utils import TransformBackends, convert_data_type, convert_to_dst_type, get_equivalent_dtype from scipy import stats from torch import Tensor @@ -306,10 +306,12 @@ def prepare_maisi_controlnet_json_dataloader( LoadImaged(keys=["image", "label"], image_only=True, ensure_channel_first=True), Orientationd(keys=["label"], axcodes="RAS"), EnsureTyped(keys=["label"], dtype=torch.uint8, track_meta=True), - Lambdad(keys="top_region_index", func=lambda x: torch.FloatTensor(x)), - Lambdad(keys="bottom_region_index", func=lambda x: torch.FloatTensor(x)), + Lambdad(keys="top_region_index", func=lambda x: torch.FloatTensor(x), allow_missing_keys=True), + Lambdad(keys="bottom_region_index", func=lambda x: torch.FloatTensor(x), allow_missing_keys=True), Lambdad(keys="spacing", func=lambda x: torch.FloatTensor(x)), - Lambdad(keys=["top_region_index", "bottom_region_index", "spacing"], func=lambda x: x * 1e2), + Lambdad( + keys=["top_region_index", "bottom_region_index", "spacing"], func=lambda x: x * 1e2, allow_missing_keys=True + ), ] train_transforms, val_transforms = Compose(common_transform), Compose(common_transform) @@ -706,7 +708,20 @@ def dynamic_infer(inferer, model, images): Returns: torch.Tensor: The output from the model or the inferer, depending on the input size. """ - if torch.numel(images[0:1, 0:1, ...]) < math.prod(inferer.roi_size): + if torch.numel(images[0:1, 0:1, ...]) <= math.prod(inferer.roi_size): return model(images) else: - return inferer(network=model, inputs=images) + # Extract the spatial dimensions from the images tensor (H, W, D) + spatial_dims = images.shape[2:] + orig_roi = inferer.roi_size + + # Check that roi has the same number of dimensions as spatial_dims + if len(orig_roi) != len(spatial_dims): + raise ValueError(f"ROI length ({len(orig_roi)}) does not match spatial dimensions ({len(spatial_dims)}).") + + # Iterate and adjust each ROI dimension + adjusted_roi = [min(roi_dim, img_dim) for roi_dim, img_dim in zip(orig_roi, spatial_dims)] + inferer.roi_size = adjusted_roi + output = inferer(network=model, inputs=images) + inferer.roi_size = orig_roi + return output