{"success":true,"data":{"author":"Jun Cheng et al., Southern Medical University, Guangzhou (converted by MedOtter)","bucketAccessType":"public","bucketName":"MedOtter/Figshare_Brain_Tumor","bucketPrefix":"","category":"Health & Biomedical","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:12.277Z","description":"This is the widely used brain tumor dataset of Cheng et al., published on Figshare with their PLoS ONE 2015 paper on tumor retrieval and classification. It contains 3,064 T1-weighted contrast-enhanced MRI slices from 233 patients with three tumor types: meningioma (708 slices), glioma (1,426) and pituitary tumor (930). Every slice comes with a binary tumor mask and the original hand-delineated tumor border. The MedOtter conversion stores each slice as a 512 x 512 16-bit PNG with its mask, border polyline, patient ID, label and the five patient-disjoint cross-validation folds of the original release; fold 1 is used as test (542 slices) and folds 2 to 5 as train (2,522). As measured 2026-11-14 the data are four Parquet files totalling about 0.7 GB; load with datasets (load_dataset('MedOtter/Figshare_Brain_Tumor')) or pyarrow. A standard benchmark for brain tumor classification and segmentation with CNNs and transformers. The authors release the data on Figshare under CC BY 4.0; cite Cheng et al. 2015.","displayTitle":"Figshare brain tumor MRI dataset (Cheng et al.): meningioma, glioma and pituitary tumor","doi":"10.6084/m9.figshare.1512427.v8","id":"fef2925a-0b57-5c1d-bfbe-196f1f0437b3","license":"CC-BY-4.0","licenseEvidenceUrl":"https://figshare.com/articles/dataset/brain_tumor_dataset/1512427","objectCountEstimate":6,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1371/journal.pone.0140381","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/Figshare_Brain_Tumor\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"MedOtter/Figshare_Brain_Tumor\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=MedOtter, license:cc-by-4.0, size_categories:1K<n<10K, downloads=671"],"format":"mixed","generatedAt":"2026-11-14T14:15:18.472Z","layout":{"example":"data/test-00000-of-00001.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"50415231150415823115fc1b4c15ba061500120000c118140200000032340506003505060037010600330512043630010c04363105060518043638050c003901060037051e0038090605240438320118043833050605420038055420383603000000313332090700330907003405070034091500340915043735051500370931043737090e0038090700390507043830090700310907093f0038093f0039092a0039052a04323305460432330546043233051c043234053f043234053f043235053f003209a1043238092a0038092a00380946003905a8043239059a0432390554043330051504333005150433300d15053f043330053f043330053f0433310570043335091c0035091c0035091c0036091c0036058c043337090e0037050e0034095b0034095b0034095b0034095b0034095b0034095b0434310931003105cb043431250a043433054d043433091c003705af0034295e0034295000342950003409e704353205310435320531043532054d04353205b604353205b604353205b60035294200352942003529c00435380562043538251104353905850035293b0035293b0035293b003609d2003609d200360970003609700436320538043632054d0436360d07055404363625650436360593043636059304363605930036293b0036293b04363705bd003629d5003629d50036099300360993003609930437","encoding":"hex","key":"data/test-00000-of-00001.parquet","truncatedAt":512}],"sampleObjects":["data/test-00000-of-00001.parquet","data/train-00000-of-00003.parquet","data/train-00001-of-00003.parquet","data/train-00002-of-00003.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n  - image-segmentation\n  - image-classification\ntags:\n  - medical\n  - mri\n  - brain\n  - tumor\n  - segmentation\n  - cheng2015\nsize_categories:\n  - 1K<n<10K\n---\n\n# Figshare Brain Tumor Dataset (Cheng et al. 2015)\n\nBrain tumor MRI dataset from Cheng et al., hosted on Figshare. 3064 T1-weighted\ncontrast-enhanced MRI slices from 233 patients with three kinds of brain tumor:\n\n- Meningioma — 708 slices\n- Glioma — 1426 slices\n- Pituitary tumor — 930 slices\n\nEach slice ships with both a per-pixel binary tumor mask and the original\nhand-delineated tumor border polyline.\n\n## Schema\n\n| Column         | Type        | Description |\n|----------------|-------------|-------------|\n| `id`           | `string`    | Slice ID, \"1\" to \"3064\" |\n| `pid`          | `string`    | Patient ID (used to keep splits patient-disjoint) |\n| `label_id`     | `int32`     | 1 = meningioma, 2 = glioma, 3 = pituitary |\n| `label_name`   | `string`    | \"meningioma\" / \"glioma\" / \"pituitary\" |\n| `fold`         | `int32`     | Pre-computed 5-fold CV index (1..5, patient-disjoint) |\n| `image`        | `Image`     | 16-bit grayscale PNG, 512×512, raw T1-CE MRI |\n| `mask`         | `Image`     | 8-bit grayscale PNG, 512×512, 0 = bg / 255 = tumor |\n| `tumor_border` | `list[float]` | Polyline coordinates `[x1, y1, x2, y2, ...]` of the manual border |\n\n## Splits\n\n| Split | Source     | Count |\n|-------|------------|------:|\n| train | folds 2-5  | 2522 |\n| test  | fold 1     |  542 |\n\nThe `fold` column is preserved on every row so users can run other CV protocols\n(rotate which fold is held out, etc.). Fold assignment is patient-level, so no\npatient appears in both train and test of any fold.\n\n## Acquisition\n\nT1-weighted contrast-enhanced MRI (after Gd-DTPA injection, 0.1 mmol/kg @ 2 ml/s)\nacquired between 2005-09 and 2010-10 at Nanfang Hospital (Guangzhou) and General\nHospital of Tianjin Medical University, China. In-plane 512×512, pixel spacing\n0.49×0.49 mm, slice thickness 6 mm, slice gap 1 mm.\n\n## Mask sources\n\nThe upstream `.mat` files ship two semantically equivalent encodings of the\nsame manual annotation:\n1. `cjdata.tumorBorder` — raw polyline coordinates as drawn by the radiologist.\n2. `cjdata.tumorMask` — the convenience rasterisation of that polyline as a\n   binary 512×512 image.\n\nThis mirror exposes both: `mask` is the rasterised version (recommended ground\ntruth for segmentation), `tumor_border` is the raw polyline.\n\n## License\n\nCC BY 4.0 — free use including commercial, attribution required.\n\n## Citation\n\n```bibtex\n@article{cheng2015enhanced,\n  title   = {Enhanced Performance of Brain Tumor Classification via Tumor Region Augmentation and Partition},\n  author  = {Cheng, Jun and Huang, Wei and Cao, Shuangliang and Yang, Ru and Yang, Wei and Yun, Zhaoqiang and Wang, Zhijian and Feng, Qianjin},\n  journal = {PLOS ONE},\n  volume  = {10},\n  number  = {10},\n  pages   = {e0140381},\n  year    = {2015},\n  doi     = {10.1371/journal.pone.0140381}\n}\n```\n\n## Source\n\n- Figshare: <https://figshare.com/articles/dataset/brain_tumor_dataset/1512427>\n- DOI: 10.6084/m9.figshare.1512427\n- Paper: <https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0140381>\n","encoding":"utf8","key":"README.md","name":"README.md","size":3228,"truncated":false}],"sourceObjectKey":"data/test-00000-of-00001.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/Figshare_Brain_Tumor\")\nprint(ds)","starterSamplePath":"data/test-00000-of-00001.parquet","totalObjects":6},"summary":"This is the widely used brain tumor dataset of Cheng et al., published on Figshare with their PLoS ONE 2015 paper on tumor retrieval and classification.","title":"Figshare brain tumor MRI dataset (Cheng et al.): meningioma, glioma and pituitary tumor","updatedAt":"2026-11-14T17:25:38.517Z","visibility":"public","websiteUrl":"https://figshare.com/articles/dataset/brain_tumor_dataset/1512427"}}