Add dataset card

README.md

@@ -0,0 +1,381 @@
+---
+license: cc-by-sa-4.0
+task_categories:
+- image-segmentation
+language:
+- en
+tags:
+- medical
+- image
+- ct
+- mri
+- multi-organ
+- tumor
+- segmentation
+- detection
+pretty_name: 'msd-lite'
+size_categories:
+- n<1K
+---
+
+
+## About
+This is a preprocessed redistribution of [Medical Segmentation Decathlon](http://medicaldecathlon.com/dataaws/), which is released under the `CC BY-SA 4.0` license.
+
+**Dataset summary:** 1,741 annotated 3D volumes across the 10 Medical Segmentation Decathlon tasks (brain, heart, liver, hippocampus, prostate, lung, pancreas, hepatic vessel, spleen, colon), each with an organ and/or tumour mask.
+
+**Contents of this repository:**
+
+- `MSD-BrainTumour/Images-FLAIR/` — 484 files
+- `MSD-BrainTumour/Images-T1w/` — 484 files
+- `MSD-BrainTumour/Images-T1gd/` — 484 files
+- `MSD-BrainTumour/Images-T2w/` — 484 files
+- `MSD-BrainTumour/Masks/` — 484 files
+- `MSD-Colon/Images/` — 126 files
+- `MSD-Colon/Masks/` — 126 files
+- `MSD-Heart/Images/` — 20 files
+- `MSD-Heart/Masks/` — 20 files
+- `MSD-HepaticVessel/Images/` — 303 files
+- `MSD-HepaticVessel/Masks/` — 303 files
+- `MSD-Hippocampus/Images/` — 260 files
+- `MSD-Hippocampus/Masks/` — 260 files
+- `MSD-Liver/Images/` — 131 files
+- `MSD-Liver/Masks/` — 131 files
+- `MSD-Lung/Images/` — 63 files
+- `MSD-Lung/Masks/` — 63 files
+- `MSD-Pancreas/Images/` — 281 files
+- `MSD-Pancreas/Masks/` — 281 files
+- `MSD-Prostate/Images-T2w/` — 32 files
+- `MSD-Prostate/Images-ADC/` — 32 files
+- `MSD-Prostate/Masks/` — 32 files
+- `MSD-Spleen/Images/` — 41 files
+- `MSD-Spleen/Masks/` — 41 files
+
+📝 Landmark annotations, visualization figures and the benchmark plan files live in 🔥[MedVision](https://huggingface.co/datasets/YongchengYAO/MedVision)🔥, where you can load the complete images and annotations from dataset configs.
+
+
+## Relation to the source dataset
+
+| | |
+| --- | --- |
+| In the source | 10 task archives, each shipping a labelled training set (`imagesTr`/`labelsTr`) and an unlabelled test set (`imagesTs`); BrainTumour images are 4-channel and Prostate images 2-channel 4D NIfTI |
+| Excluded here | the unlabelled `imagesTs` volumes of all 10 tasks (no public masks), and the per-task `dataset.json` |
+| **In this repo** | **484 `MSD-BrainTumour/Images-FLAIR` + 484 `MSD-BrainTumour/Images-T1w` + 484 `MSD-BrainTumour/Images-T1gd` + 484 `MSD-BrainTumour/Images-T2w` + 484 `MSD-BrainTumour/Masks` + 126 `MSD-Colon/Images` + 126 `MSD-Colon/Masks` + 20 `MSD-Heart/Images` + 20 `MSD-Heart/Masks` + 303 `MSD-HepaticVessel/Images` + 303 `MSD-HepaticVessel/Masks` + 260 `MSD-Hippocampus/Images` + 260 `MSD-Hippocampus/Masks` + 131 `MSD-Liver/Images` + 131 `MSD-Liver/Masks` + 63 `MSD-Lung/Images` + 63 `MSD-Lung/Masks` + 281 `MSD-Pancreas/Images` + 281 `MSD-Pancreas/Masks` + 32 `MSD-Prostate/Images-T2w` + 32 `MSD-Prostate/Images-ADC` + 32 `MSD-Prostate/Masks` + 41 `MSD-Spleen/Images` + 41 `MSD-Spleen/Masks`** |
+
+**Every labelled case of all 10 tasks is included - 1,741 cases.** No format conversion was required: the source already ships `nii.gz`, and image and mask voxel data are carried over unmodified. What is derived here is the RAS+ reorientation, the `MSD-<Task>/{Images,Masks}` layout, and the **splitting of the 4D multi-channel volumes into one 3D volume per channel**.
+
+Because of that split, two tasks have **no plain `Images/` folder**: `MSD-BrainTumour` ships `Images-FLAIR`, `Images-T1w`, `Images-T1gd`, `Images-T2w` and `MSD-Prostate` ships `Images-T2w`, `Images-ADC`. Each of those tasks still has a single `Masks/` folder, which pairs case-for-case with every one of its modality folders.
+
+**Why `-Lite`?** The suffix marks this as a *derived* redistribution rather than a copy of the source. These are **preprocessed** volumes — every case has been format-converted where needed, geometry-normalised and reoriented to RAS+ — and for some sources cases or modalities are excluded as well (see the table above). Use it to reproduce MedVision, not as a substitute for the original release. See [Preprocessing](#preprocessing) below for exactly what was changed.
+
+
+## Preprocessing
+
+- Built from the 10 official task archives (`Task01_BrainTumour.tar` ... `Task10_Colon.tar`) at `https://msd-for-monai.s3-us-west-2.amazonaws.com`.
+
+- `Task0N_<X>/` is renamed to `MSD-<X>/`, `imagesTr/` to `Images/` and `labelsTr/` to `Masks/`; the unlabelled `imagesTs/` and the per-task `dataset.json` are dropped.
+
+- 4D multi-channel images are split into one 3D volume per channel: BrainTumour into `Images-FLAIR` / `Images-T1w` / `Images-T1gd` / `Images-T2w` (channel order as published), Prostate into `Images-T2w` / `Images-ADC`. The mask is not duplicated per channel.
+
+- macOS resource forks and metadata from the source archives (`._*`, `.DS_Store`) are removed.
+
+- No format conversion and no resampling - labels are the source's. Only the orientation is standardized to RAS+.
+
+📝 **ShareAlike.** MSD is `CC BY-SA 4.0`, so this mirror is redistributed under the same licence, the modifications made are stated above, and **no additional restrictions are imposed** - the repository is public and ungated, and any redistribution of it or of a derivative must stay that way. Note also that several MSD tasks are re-cropped by other public collections (e.g. ULS23 re-uses MSD lesions), so combining MSD with such a dataset can place the same lesion on both sides of a train/test split.
+
+
+## Segmentation Labels
+
+```python
+labels_map_MSD_BrainTumour = {
+    "1": "edema of brain",
+    "2": "non-enhancing brain tumor",
+    "3": "enhancing brain tumor"
+}
+labels_map_MSD_Colon = {
+    "1": "colon cancer primaries"
+}
+labels_map_MSD_HepaticVessel = {
+    "1": "liver vessel",
+    "2": "liver tumor"
+}
+labels_map_MSD_Liver = {
+    "1": "liver",
+    "2": "liver cancer"
+}
+labels_map_MSD_Lung = {
+    "1": "lung cancer"
+}
+labels_map_MSD_Pancreas = {
+    "1": "pancreas",
+    "2": "pancreas cancer"
+}
+labels_map_MSD_Heart = {
+    "1": "left atrium of heart"
+}
+labels_map_MSD_Hippocampus = {
+    "1": "anterior hippocampus",
+    "2": "posterior hippocampus"
+}
+labels_map_MSD_Prostate = {
+    "1": "peripheral zone of prostate",
+    "2": "transition zone of prostate"
+}
+labels_map_MSD_Spleen = {
+    "1": "spleen"
+}
+```
+
+
+## Landmarks
+
+```python
+landmarks_map = {
+    "P1": "most right/anterior/superior endpoint of the major axis",
+    "P2": "most left/superior/inferior endpoint of the major axis",
+    "P3": "most right/anterior/superior endpoint of the minor axis",
+    "P4": "most left/superior/inferior endpoint of the minor axis"
+}
+```
+
+
+## News
+- [2 Aug, 2026] Initial release. This dataset is integrated into 🔥[MedVision](https://huggingface.co/datasets/YongchengYAO/MedVision)🔥, where you can use these config names to load data in python:
+
+  - `MSD_BoxSize_Task01_Axial_Test`
+  - `MSD_BoxSize_Task01_Axial_Train`
+  - `MSD_BoxSize_Task01_Coronal_Test`
+  - `MSD_BoxSize_Task01_Coronal_Train`
+  - `MSD_BoxSize_Task01_Sagittal_Test`
+  - `MSD_BoxSize_Task01_Sagittal_Train`
+  - `MSD_BoxSize_Task02_Axial_Test`
+  - `MSD_BoxSize_Task02_Axial_Train`
+  - `MSD_BoxSize_Task02_Coronal_Test`
+  - `MSD_BoxSize_Task02_Coronal_Train`
+  - `MSD_BoxSize_Task02_Sagittal_Test`
+  - `MSD_BoxSize_Task02_Sagittal_Train`
+  - `MSD_BoxSize_Task03_Axial_Test`
+  - `MSD_BoxSize_Task03_Axial_Train`
+  - `MSD_BoxSize_Task03_Coronal_Test`
+  - `MSD_BoxSize_Task03_Coronal_Train`
+  - `MSD_BoxSize_Task03_Sagittal_Test`
+  - `MSD_BoxSize_Task03_Sagittal_Train`
+  - `MSD_BoxSize_Task04_Axial_Test`
+  - `MSD_BoxSize_Task04_Axial_Train`
+  - `MSD_BoxSize_Task04_Coronal_Test`
+  - `MSD_BoxSize_Task04_Coronal_Train`
+  - `MSD_BoxSize_Task04_Sagittal_Test`
+  - `MSD_BoxSize_Task04_Sagittal_Train`
+  - `MSD_BoxSize_Task05_Axial_Test`
+  - `MSD_BoxSize_Task05_Axial_Train`
+  - `MSD_BoxSize_Task05_Coronal_Test`
+  - `MSD_BoxSize_Task05_Coronal_Train`
+  - `MSD_BoxSize_Task05_Sagittal_Test`
+  - `MSD_BoxSize_Task05_Sagittal_Train`
+  - `MSD_BoxSize_Task06_Axial_Test`
+  - `MSD_BoxSize_Task06_Axial_Train`
+  - `MSD_BoxSize_Task06_Coronal_Test`
+  - `MSD_BoxSize_Task06_Coronal_Train`
+  - `MSD_BoxSize_Task06_Sagittal_Test`
+  - `MSD_BoxSize_Task06_Sagittal_Train`
+  - `MSD_BoxSize_Task07_Axial_Test`
+  - `MSD_BoxSize_Task07_Axial_Train`
+  - `MSD_BoxSize_Task07_Coronal_Test`
+  - `MSD_BoxSize_Task07_Coronal_Train`
+  - `MSD_BoxSize_Task07_Sagittal_Test`
+  - `MSD_BoxSize_Task07_Sagittal_Train`
+  - `MSD_BoxSize_Task08_Axial_Test`
+  - `MSD_BoxSize_Task08_Axial_Train`
+  - `MSD_BoxSize_Task08_Coronal_Test`
+  - `MSD_BoxSize_Task08_Coronal_Train`
+  - `MSD_BoxSize_Task08_Sagittal_Test`
+  - `MSD_BoxSize_Task08_Sagittal_Train`
+  - `MSD_BoxSize_Task09_Axial_Test`
+  - `MSD_BoxSize_Task09_Axial_Train`
+  - `MSD_BoxSize_Task09_Coronal_Test`
+  - `MSD_BoxSize_Task09_Coronal_Train`
+  - `MSD_BoxSize_Task09_Sagittal_Test`
+  - `MSD_BoxSize_Task09_Sagittal_Train`
+  - `MSD_BoxSize_Task10_Axial_Test`
+  - `MSD_BoxSize_Task10_Axial_Train`
+  - `MSD_BoxSize_Task10_Coronal_Test`
+  - `MSD_BoxSize_Task10_Coronal_Train`
+  - `MSD_BoxSize_Task10_Sagittal_Test`
+  - `MSD_BoxSize_Task10_Sagittal_Train`
+  - `MSD_BoxSize_Task11_Axial_Test`
+  - `MSD_BoxSize_Task11_Axial_Train`
+  - `MSD_BoxSize_Task11_Coronal_Test`
+  - `MSD_BoxSize_Task11_Coronal_Train`
+  - `MSD_BoxSize_Task11_Sagittal_Test`
+  - `MSD_BoxSize_Task11_Sagittal_Train`
+  - `MSD_BoxSize_Task12_Axial_Test`
+  - `MSD_BoxSize_Task12_Axial_Train`
+  - `MSD_BoxSize_Task12_Coronal_Test`
+  - `MSD_BoxSize_Task12_Coronal_Train`
+  - `MSD_BoxSize_Task12_Sagittal_Test`
+  - `MSD_BoxSize_Task12_Sagittal_Train`
+  - `MSD_BoxSize_Task13_Axial_Test`
+  - `MSD_BoxSize_Task13_Axial_Train`
+  - `MSD_BoxSize_Task13_Coronal_Test`
+  - `MSD_BoxSize_Task13_Coronal_Train`
+  - `MSD_BoxSize_Task13_Sagittal_Test`
+  - `MSD_BoxSize_Task13_Sagittal_Train`
+  - `MSD_BoxSize_Task14_Axial_Test`
+  - `MSD_BoxSize_Task14_Axial_Train`
+  - `MSD_BoxSize_Task14_Coronal_Test`
+  - `MSD_BoxSize_Task14_Coronal_Train`
+  - `MSD_BoxSize_Task14_Sagittal_Test`
+  - `MSD_BoxSize_Task14_Sagittal_Train`
+  - `MSD_MaskSize_Task01_Axial_Test`
+  - `MSD_MaskSize_Task01_Axial_Train`
+  - `MSD_MaskSize_Task01_Coronal_Test`
+  - `MSD_MaskSize_Task01_Coronal_Train`
+  - `MSD_MaskSize_Task01_Sagittal_Test`
+  - `MSD_MaskSize_Task01_Sagittal_Train`
+  - `MSD_MaskSize_Task02_Axial_Test`
+  - `MSD_MaskSize_Task02_Axial_Train`
+  - `MSD_MaskSize_Task02_Coronal_Test`
+  - `MSD_MaskSize_Task02_Coronal_Train`
+  - `MSD_MaskSize_Task02_Sagittal_Test`
+  - `MSD_MaskSize_Task02_Sagittal_Train`
+  - `MSD_MaskSize_Task03_Axial_Test`
+  - `MSD_MaskSize_Task03_Axial_Train`
+  - `MSD_MaskSize_Task03_Coronal_Test`
+  - `MSD_MaskSize_Task03_Coronal_Train`
+  - `MSD_MaskSize_Task03_Sagittal_Test`
+  - `MSD_MaskSize_Task03_Sagittal_Train`
+  - `MSD_MaskSize_Task04_Axial_Test`
+  - `MSD_MaskSize_Task04_Axial_Train`
+  - `MSD_MaskSize_Task04_Coronal_Test`
+  - `MSD_MaskSize_Task04_Coronal_Train`
+  - `MSD_MaskSize_Task04_Sagittal_Test`
+  - `MSD_MaskSize_Task04_Sagittal_Train`
+  - `MSD_MaskSize_Task05_Axial_Test`
+  - `MSD_MaskSize_Task05_Axial_Train`
+  - `MSD_MaskSize_Task05_Coronal_Test`
+  - `MSD_MaskSize_Task05_Coronal_Train`
+  - `MSD_MaskSize_Task05_Sagittal_Test`
+  - `MSD_MaskSize_Task05_Sagittal_Train`
+  - `MSD_MaskSize_Task06_Axial_Test`
+  - `MSD_MaskSize_Task06_Axial_Train`
+  - `MSD_MaskSize_Task06_Coronal_Test`
+  - `MSD_MaskSize_Task06_Coronal_Train`
+  - `MSD_MaskSize_Task06_Sagittal_Test`
+  - `MSD_MaskSize_Task06_Sagittal_Train`
+  - `MSD_MaskSize_Task07_Axial_Test`
+  - `MSD_MaskSize_Task07_Axial_Train`
+  - `MSD_MaskSize_Task07_Coronal_Test`
+  - `MSD_MaskSize_Task07_Coronal_Train`
+  - `MSD_MaskSize_Task07_Sagittal_Test`
+  - `MSD_MaskSize_Task07_Sagittal_Train`
+  - `MSD_MaskSize_Task08_Axial_Test`
+  - `MSD_MaskSize_Task08_Axial_Train`
+  - `MSD_MaskSize_Task08_Coronal_Test`
+  - `MSD_MaskSize_Task08_Coronal_Train`
+  - `MSD_MaskSize_Task08_Sagittal_Test`
+  - `MSD_MaskSize_Task08_Sagittal_Train`
+  - `MSD_MaskSize_Task09_Axial_Test`
+  - `MSD_MaskSize_Task09_Axial_Train`
+  - `MSD_MaskSize_Task09_Coronal_Test`
+  - `MSD_MaskSize_Task09_Coronal_Train`
+  - `MSD_MaskSize_Task09_Sagittal_Test`
+  - `MSD_MaskSize_Task09_Sagittal_Train`
+  - `MSD_MaskSize_Task10_Axial_Test`
+  - `MSD_MaskSize_Task10_Axial_Train`
+  - `MSD_MaskSize_Task10_Coronal_Test`
+  - `MSD_MaskSize_Task10_Coronal_Train`
+  - `MSD_MaskSize_Task10_Sagittal_Test`
+  - `MSD_MaskSize_Task10_Sagittal_Train`
+  - `MSD_MaskSize_Task11_Axial_Test`
+  - `MSD_MaskSize_Task11_Axial_Train`
+  - `MSD_MaskSize_Task11_Coronal_Test`
+  - `MSD_MaskSize_Task11_Coronal_Train`
+  - `MSD_MaskSize_Task11_Sagittal_Test`
+  - `MSD_MaskSize_Task11_Sagittal_Train`
+  - `MSD_MaskSize_Task12_Axial_Test`
+  - `MSD_MaskSize_Task12_Axial_Train`
+  - `MSD_MaskSize_Task12_Coronal_Test`
+  - `MSD_MaskSize_Task12_Coronal_Train`
+  - `MSD_MaskSize_Task12_Sagittal_Test`
+  - `MSD_MaskSize_Task12_Sagittal_Train`
+  - `MSD_MaskSize_Task13_Axial_Test`
+  - `MSD_MaskSize_Task13_Axial_Train`
+  - `MSD_MaskSize_Task13_Coronal_Test`
+  - `MSD_MaskSize_Task13_Coronal_Train`
+  - `MSD_MaskSize_Task13_Sagittal_Test`
+  - `MSD_MaskSize_Task13_Sagittal_Train`
+  - `MSD_MaskSize_Task14_Axial_Test`
+  - `MSD_MaskSize_Task14_Axial_Train`
+  - `MSD_MaskSize_Task14_Coronal_Test`
+  - `MSD_MaskSize_Task14_Coronal_Train`
+  - `MSD_MaskSize_Task14_Sagittal_Test`
+  - `MSD_MaskSize_Task14_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task01_Axial_Test`
+  - `MSD_TumorLesionSize_Task01_Axial_Train`
+  - `MSD_TumorLesionSize_Task01_Coronal_Test`
+  - `MSD_TumorLesionSize_Task01_Coronal_Train`
+  - `MSD_TumorLesionSize_Task01_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task01_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task02_Axial_Test`
+  - `MSD_TumorLesionSize_Task02_Axial_Train`
+  - `MSD_TumorLesionSize_Task02_Coronal_Test`
+  - `MSD_TumorLesionSize_Task02_Coronal_Train`
+  - `MSD_TumorLesionSize_Task02_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task02_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task03_Axial_Test`
+  - `MSD_TumorLesionSize_Task03_Axial_Train`
+  - `MSD_TumorLesionSize_Task03_Coronal_Test`
+  - `MSD_TumorLesionSize_Task03_Coronal_Train`
+  - `MSD_TumorLesionSize_Task03_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task03_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task04_Axial_Test`
+  - `MSD_TumorLesionSize_Task04_Axial_Train`
+  - `MSD_TumorLesionSize_Task04_Coronal_Test`
+  - `MSD_TumorLesionSize_Task04_Coronal_Train`
+  - `MSD_TumorLesionSize_Task04_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task04_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task05_Axial_Test`
+  - `MSD_TumorLesionSize_Task05_Axial_Train`
+  - `MSD_TumorLesionSize_Task05_Coronal_Test`
+  - `MSD_TumorLesionSize_Task05_Coronal_Train`
+  - `MSD_TumorLesionSize_Task05_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task05_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task06_Axial_Test`
+  - `MSD_TumorLesionSize_Task06_Axial_Train`
+  - `MSD_TumorLesionSize_Task06_Coronal_Test`
+  - `MSD_TumorLesionSize_Task06_Coronal_Train`
+  - `MSD_TumorLesionSize_Task06_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task06_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task07_Axial_Test`
+  - `MSD_TumorLesionSize_Task07_Axial_Train`
+  - `MSD_TumorLesionSize_Task07_Coronal_Test`
+  - `MSD_TumorLesionSize_Task07_Coronal_Train`
+  - `MSD_TumorLesionSize_Task07_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task07_Sagittal_Train`
+  - `MSD_TumorLesionSize_Task08_Axial_Test`
+  - `MSD_TumorLesionSize_Task08_Axial_Train`
+  - `MSD_TumorLesionSize_Task08_Coronal_Test`
+  - `MSD_TumorLesionSize_Task08_Coronal_Train`
+  - `MSD_TumorLesionSize_Task08_Sagittal_Test`
+  - `MSD_TumorLesionSize_Task08_Sagittal_Train`
+
+
+## Data Usage Agreement
+By using the dataset, you agree to the terms as follow.
+- You must comply with the original `CC BY-SA 4.0` license terms of the source dataset.
+- You are recommended to refer to the source of this dataset in any publication: `https://huggingface.co/datasets/YongchengYAO/MSD-Lite`
+- You must cite the original publication(s):
+  - https://doi.org/10.1038/s41467-022-30695-9
+
+
+## Official Release
+For more information, please go to the official site: http://medicaldecathlon.com/dataaws/
+
+
+## Download from Huggingface
+```python
+# python
+from huggingface_hub import snapshot_download
+snapshot_download(repo_id="YongchengYAO/MSD-Lite", repo_type='dataset', local_dir="/your/local/folder")
+```
+