[feat] Add data_prep and dummy_model configurations for two BraTS 2024 challenges

[vpchung]

BraTS 2024 features two new challenges that requires new data_prep and dummy_model MLCubes:

• BraTS-MEN-RT (Meningioma Radiotherapy)
• BraTS-Path (Pathology)

Specifically, BraTS-MEN-RT will only contain a single modality for the data (t1c) and will use gtv as the labels; BraTS-Path will contain a single directory of PNG files as the data and will use a single 2-column CSV as the labels.

Changelog

• update prepare.py to handle the raw data of BraTS-MEN-RT and BraTS-Path
• update sanity_check.py to assert the expectations mentioned above ^
• update infer.py to output a 2-column CSV for BraTS-Path (minor changes needed for BraTS-MEN-RT, e.g. using underscore instead of a dash in the filenames)
• add pandas and numpy to dummy model requirements
• update mlcube.yaml files
• add project code for mock metrics (for the compatibility test)
• add project code for mock data prep (for the compatibility test)

Preview

BraTS-MEN-RT

Prepare the raw data

$ python data_prep/project/mlcube.py prepare \
  --data_path ~/brats2024-sample-data/BraTS-MEN-RT/ \
  --labels_path ~/brats2024-sample-data/BraTS-MEN-RT/ \
  --output_path test_data \
  --output_labels_path test_labels \
  --parameters_file data_prep/mlcube/workspace/parameters-rt.yaml 
$
$ tree test_*
test_data
├── BraTS-MEN-RT-xxxx-x
│   └── BraTS-MEN-RT-xxxx-x_t1c.nii.gz
├── BraTS-MEN-RT-yyyy-y
│   └── BraTS-MEN-RT-yyyy-y_t1c.nii.gz
└── BraTS-MEN-RT-zzzz-z
    └── BraTS-MEN-RT-zzzz-z_t1c.nii.gz
test_labels
├── BraTS-MEN-RT-xxxx-x_gtv.nii.gz
├── BraTS-MEN-RT-yyyy-y_gtv.nii.gz
└── BraTS-MEN-RT-zzzz-z_gtv.nii.gz

Validate the prepared data

$ python data_prep/project/mlcube.py sanity_check \
  --data_path test_data/ \
  --labels_path test_labels/ \
  --parameters_file data_prep/mlcube/workspace/parameters-rt.yaml 
$

Run dummy model

$ python dummy_model/project/mlcube.py infer \
  --data_path test_data/ \
  --parameters_file dummy_model/mlcube/workspace/parameters-rt.yaml \
  --output_path test_predictions/
$
$ tree test_predictions
test_predictions
├── BraTS-MEN-RT-xxxx-x.nii.gz
├── BraTS-MEN-RT-yyyy-y.nii.gz
└── BraTS-MEN-RT-zzzz-z.nii.gz

BraTS-Path

Prepare the raw data

$ python data_prep/project/mlcube.py prepare \
  --data_path ~/brats2024-sample-data/BraTS-Path/ \
  --labels_path ~/brats2024-sample-data/BraTS-Path/ \
  --output_path test_data \
  --output_labels_path test_labels \
  --parameters_file data_prep/mlcube/workspace/parameters-path.yaml 
$
$ tree test_*
test_data
├── BraTSPath_cohort_xxxxxxx.png
├── BraTSPath_cohort_yyyyyyy.png
└── BraTSPath_cohort_zzzzzzz.png
test_labels
└── BraTS-PATH-cohort-Labels.csv

Validate the prepared data

$ python data_prep/project/mlcube.py sanity_check \
  --data_path test_data/ \
  --labels_path test_labels/ \
  --parameters_file data_prep/mlcube/workspace/parameters-path.yaml 
$

Run dummy model

$ python dummy_model/project/mlcube.py infer \
  --data_path test_data/ \
  --parameters_file dummy_model/mlcube/workspace/parameters-path.yaml \
  --output_path test_predictions/
$
$ cat test_predictions/predictions.csv
SubjectID,Prediction
BraTSPath_cohort_xxxxxxx.png,A
BraTSPath_cohort_yyyyyyy.png,B
BraTSPath_cohort_zzzzzzz.png,C

where A, B, and C are integers from 0 to 5.

[github-actions]

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅