YongchengYAO/AMOS22-Lite

About

This is a preprocessed redistribution of AMOS22 (Zenodo), which is released under the CC BY 4.0 license.

Dataset summary: 360 abdominal scans (300 CT + 60 MRI) with 15-class multi-organ segmentation masks.

Contents of this repository:

• AMOS22-CT/Images/ — 300 files
• AMOS22-CT/Masks/ — 300 files
• AMOS22-MRI/Images/ — 60 files
• AMOS22-MRI/Masks/ — 60 files

📝 Landmark annotations, visualization figures and the benchmark plan files live in 🔥MedVision🔥, where you can load the complete images and annotations from dataset configs.

Relation to the source dataset

In the source	the AMOS 2022 release - abdominal CT and MRI across train/val/test; expert masks are public for the train+val split only (300 CT + 60 MRI)
Excluded here	the unlabelled test images (imagesTs), which ship no public masks
In this repo	300 AMOS22-CT/Images + 300 AMOS22-CT/Masks + 60 AMOS22-MRI/Images + 60 AMOS22-MRI/Masks

Every labelled case is included - 300 CT and 60 MRI. No format conversion was required: the source already ships nii.gz, and the image voxel data is carried over unmodified. What is derived here is the RAS+ reorientation and the modality-split layout that replaces the source's imagesTr/imagesVa + labelsTr/labelsVa folders.

CT and MRI are shipped as separate folder pairs because MedVision treats them as two tasks, and because the source encodes the modality only in the case number (amos_0001-amos_0506 are CT, amos_0507+ are MRI) - a fact that is easy to lose once the files are pooled.

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 below for exactly what was changed.

Preprocessing

• Built from the single official archive amos22.zip (https://zenodo.org/records/7155725/files/amos22.zip).

• imagesTr + imagesVa and labelsTr + labelsVa are pooled and then split by case number into AMOS22-CT/ (amos_0001-amos_0506) and AMOS22-MRI/ (amos_0507+). Filenames keep the source stem, so a case ID is identical across Images/ and Masks/.

• No format conversion and no resampling - the source is already nii.gz, and the 15-label mask encoding is the source's. Only the orientation is standardized to RAS+.

Segmentation Labels

labels_map = {
    "1": "spleen",
    "2": "right kidney",
    "3": "left kidney",
    "4": "gall bladder",
    "5": "esophagus",
    "6": "liver",
    "7": "stomach",
    "8": "aorta",
    "9": "postcava",
    "10": "pancreas",
    "11": "right adrenal gland",
    "12": "left adrenal gland",
    "13": "duodenum",
    "14": "bladder",
    "15": "prostate/uterus"
}

News

• [2 Aug, 2026] Initial release. This dataset is integrated into 🔥MedVision🔥, where you can use these config names to load data in python:

  • AMOS22_BoxSize_Task01_Axial_Test
  • AMOS22_BoxSize_Task01_Axial_Train
  • AMOS22_BoxSize_Task01_Coronal_Test
  • AMOS22_BoxSize_Task01_Coronal_Train
  • AMOS22_BoxSize_Task01_Sagittal_Test
  • AMOS22_BoxSize_Task01_Sagittal_Train
  • AMOS22_BoxSize_Task02_Axial_Test
  • AMOS22_BoxSize_Task02_Axial_Train
  • AMOS22_BoxSize_Task02_Coronal_Test
  • AMOS22_BoxSize_Task02_Coronal_Train
  • AMOS22_BoxSize_Task02_Sagittal_Test
  • AMOS22_BoxSize_Task02_Sagittal_Train
  • AMOS22_MaskSize_Task01_Axial_Test
  • AMOS22_MaskSize_Task01_Axial_Train
  • AMOS22_MaskSize_Task01_Coronal_Test
  • AMOS22_MaskSize_Task01_Coronal_Train
  • AMOS22_MaskSize_Task01_Sagittal_Test
  • AMOS22_MaskSize_Task01_Sagittal_Train
  • AMOS22_MaskSize_Task02_Axial_Test
  • AMOS22_MaskSize_Task02_Axial_Train
  • AMOS22_MaskSize_Task02_Coronal_Test
  • AMOS22_MaskSize_Task02_Coronal_Train
  • AMOS22_MaskSize_Task02_Sagittal_Test
  • AMOS22_MaskSize_Task02_Sagittal_Train

Data Usage Agreement

By using the dataset, you agree to the terms as follow.

• You must comply with the original CC BY 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/AMOS22-Lite
• You must cite the original publication(s):
  • https://doi.org/10.48550/arXiv.2206.08023

Official Release

For more information, please go to the official site: https://amos22.grand-challenge.org

Download from Huggingface

# python
from huggingface_hub import snapshot_download
snapshot_download(repo_id="YongchengYAO/AMOS22-Lite", repo_type='dataset', local_dir="/your/local/folder")