Datasets:
metadata
license: cc-by-nc-sa-2.0
task_categories:
- image-segmentation
language:
- en
tags:
- medical
- image
pretty_name: totalsegmentator-mr-lite
size_categories:
- n<1K
About
This is a derivative of the TotalSegmentator dataset
- 616 MR images and corresponding segmentation mask of 50 structures
- We combined multiple segmentation masks into a single
nii.gz
file under the folderMasks
, and moved all MR images to the folderImages
. - All images and masks are renamed according to case IDs.
This dataset is released under the CC BY-NC-SA 2.0 license.
Official Release
- GitHub (official): https://github.com/wasserth/TotalSegmentator (Apache-2.0 license)
- Data (official): https://zenodo.org/records/14710732 (CC BY-NC-SA 2.0 license)
Download from Huggingface
#!/bin/bash
pip install huggingface-hub[cli]
huggingface-cli login --token $HF_TOKEN
# python
from huggingface_hub import snapshot_download
snapshot_download(repo_id="YongchengYAO/TotalSegmentator-MR-Lite", repo_type='dataset', local_dir="/your/local/folder")
Segmentation Labels
labels_MR = {
1: "spleen",
2: "kidney_right",
3: "kidney_left",
4: "gallbladder",
5: "liver",
6: "stomach",
7: "pancreas",
8: "adrenal_gland_right",
9: "adrenal_gland_left",
10: "lung_left",
11: "lung_right",
12: "esophagus",
13: "small_bowel",
14: "duodenum",
15: "colon",
16: "urinary_bladder",
17: "prostate",
18: "sacrum",
19: "vertebrae",
20: "intervertebral_discs",
21: "spinal_cord",
22: "heart",
23: "aorta",
24: "inferior_vena_cava",
25: "portal_vein_and_splenic_vein",
26: "iliac_artery_left",
27: "iliac_artery_right",
28: "iliac_vena_left",
29: "iliac_vena_right",
30: "humerus_left",
31: "humerus_right",
32: "scapula_left",
33: "scapula_right",
34: "clavicula_left",
35: "clavicula_right",
36: "femur_left",
37: "femur_right",
38: "hip_left",
39: "hip_right",
40: "gluteus_maximus_left",
41: "gluteus_maximus_right",
42: "gluteus_medius_left",
43: "gluteus_medius_right",
44: "gluteus_minimus_left",
45: "gluteus_minimus_right",
46: "autochthon_left",
47: "autochthon_right",
48: "iliopsoas_left",
49: "iliopsoas_right",
50: "brain"
}