Conver metadata types to native and assign metadata to in-mem output image

[MMelQin]

This is to address issue Attach meta data to the image object which is output by the segmentation operator

Please note, this PR also implements a function to convert the data type of a few known metadata attributes of the input Image object, as a workaround for the newly raised issue Represent DICOM metadata attribute with built-in or App SDK types

Example output of the converted metadata dict within the inference operator log:

Executing operator LiverTumorSegOperator (Process ID: 31824, Operator ID: 7beb6929-5a19-462e-9896-4072c7baf217)
Operator output folder path: /home/mqin/src/monai-app-sdk/examples/apps/.monai_workdir/operators/7beb6929-5a19-462e-9896-4072c7baf217/0/output/saved_images_folder
Converted Image object metadata:
series_instance_uid: 1.2.826.0.1.3680043.2.1125.1.68102559796966796813942775094416763, type <class 'str'>
modality: CT, type <class 'str'>
series_description: No series description, type <class 'str'>
row_pixel_spacing: 1.0, type <class 'float'>
col_pixel_spacing: 1.0, type <class 'float'>
depth_pixel_spacing: 1.0, type <class 'float'>
row_direction_cosine: [-1.0, 0.0, 0.0], type <class 'list'>
col_direction_cosine: [0.0, -1.0, 0.0], type <class 'list'>
depth_direction_cosine: [0.0, 0.0, 1.0], type <class 'list'>
dicom_affine_transform: [[-1.  0.  0.  0.]
 [ 0. -1.  0.  0.]
 [ 0.  0.  1.  0.]
 [ 0.  0.  0.  1.]], type <class 'numpy.ndarray'>
nifti_affine_transform: [[ 1. -0. -0. -0.]
 [-0.  1. -0. -0.]
 [ 0.  0.  1.  0.]
 [ 0.  0.  0.  1.]], type <class 'numpy.ndarray'>
file written: /home/mqin/src/monai-app-sdk/examples/apps/.monai_workdir/operators/7beb6929-5a19-462e-9896-4072c7baf217/0/output/saved_images_folder/1.2.826.0.1.3680043.2.1125.1/1.2.826.0.1.3680043.2.1125.1_seg.nii.gz.
file written: /home/mqin/src/monai-app-sdk/examples/apps/.monai_workdir/operators/7beb6929-5a19-462e-9896-4072c7baf217/0/output/saved_images_folder/1.2.826.0.1.3680043.2.1125.1/1.2.826.0.1.3680043.2.1125.1.nii.gz.
Output Seg image numpy array shaped: (515, 440, 440)
Output Seg image pixel max value: 2

[codecov]

Codecov Report

Merging #186 (a01f327) into main (9006298) will decrease coverage by 0.18%.
The diff coverage is 4.00%.

@@            Coverage Diff             @@
##             main     #186      +/-   ##
==========================================
- Coverage   35.29%   35.10%   -0.19%     
==========================================
  Files          64       64              
  Lines        3077     3096      +19     
==========================================
+ Hits         1086     1087       +1     
- Misses       1991     2009      +18     

Impacted Files	Coverage Δ
...i/deploy/operators/monai_seg_inference_operator.py	28.17% <4.00%> (-2.69%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9006298...a01f327. Read the comment docs.

[AndreasHeumann]

Looks good to me, just a minor issue.

[AndreasHeumann]

Do we need this str() cast here since when the uid had already been converted to str?

[MMelQin]

True, unnecessary, I left it there just for double safety and be explicit.