{"success":true,"data":{"author":"Pengbo Liu, Hu Han et al.; mirrored by MedOtter","bucketAccessType":"public","bucketName":"MedOtter/CTPelvic1K","bucketPrefix":"","category":"Health & Biomedical","cloudProvider":"huggingface","createdAt":"2026-11-17T19:29:33.567Z","description":"Pelvic bone CT with dense segmentations of the sacrum, left hip, right hip and lumbar vertebrae, from the CTPelvic1K dataset. This mirror carries only the two subsets whose images the authors released under their own license: dataset6 CLINIC (103 clinical CT scans) and dataset7 CLINIC-metal (14 scans with metal implants and their artifacts), each with masks, as NIfTI in CTPelvic1K_dataset6_data/, CTPelvic1K_dataset6_Anonymized_mask/, CTPelvic1K_dataset7_data/ and CTPelvic1K_dataset7_mask/, indexed by train.jsonl. Images for datasets 1 to 5 (drawn from KiTS19, BTCV, COLONOG and MSD under their own terms) are intentionally not included. About 10 GB (measured 2026-11-17). Use for pelvic bone segmentation and for testing robustness to metal artifacts. Load with nibabel or SimpleITK. Mirrored on Hugging Face by MedOtter from the authors' release, with a card documenting provenance and any normalization. The authors' Zenodo record 4588403 (annotations and the CLINIC and CLINIC-metal data) is CC BY 4.0; cite Liu, Han et al., CTPelvic1K.","displayTitle":"CTPelvic1K CLINIC and CLINIC-metal: pelvic CT with sacrum, hip and lumbar vertebra segmentations (117 cases)","doi":"10.5281/zenodo.4588403","huggingFaceUrl":"https://huggingface.co/datasets/MedOtter/CTPelvic1K","id":"bd97b765-30b1-5e00-9931-6278e4331c35","license":"CC-BY-4.0","licenseEvidenceUrl":"https://zenodo.org/records/4588403","objectCountEstimate":237,"probeStatus":"ok","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/CTPelvic1K\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"MedOtter/CTPelvic1K\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=MedOtter, license:cc-by-4.0, size_categories:n<1K, format:json, downloads=1820"],"format":"mixed","generatedAt":"2026-11-17T19:40:34.343Z","layout":{"example":"train.jsonl","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"{\"image\": \"data/nii/CTPelvic1K/CTPelvic1K_dataset6_data/dataset6_CLINIC_0001_data.nii.gz\", \"mask\": \"data/nii/CTPelvic1K/CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0001_mask_4label.nii.gz\", \"label\": [\"sacrum\", \"left hip\", \"right hip\", \"lumbar vertebrae\", \"hip\"], \"modality\": \"CT\", \"dataset\": \"CTPelvic1K\", \"official_split\": \"train\", \"patient_id\": \"dataset6_CLINIC_0001_data.nii.gz\"}\n{\"image\": \"data/nii/CTPelvic1K/CTPelvic1K_dataset6_data/dataset6_CLINIC_0002_data.nii.gz\", \"mask\": \"data/nii/CTPelvic1K/C","encoding":"utf8","key":"train.jsonl","truncatedAt":512}],"sampleObjects":["CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0001_mask_4label.nii.gz","CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0002_mask_4label.nii.gz","CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0003_mask_4label.nii.gz","CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0004_mask_4label.nii.gz","CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0005_mask_4label.nii.gz"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n- image-segmentation\ntags:\n- medical\n- CT\n- segmentation\n- CTPelvic1K\nsize_categories:\n- n<1K\nconfigs:\n- config_name: default\n  data_files:\n  - split: train\n    path: train.jsonl\n---\n\n# CTPelvic1K Dataset\n\n## Dataset Description\n\nThe CTPelvic1K dataset for pelvic bone segmentation from CT scans. This dataset contains CT scans with dense segmentation annotations.\n\n### Dataset Details\n\n- **Modality**: CT\n- **Target**: sacrum, left hip, right hip, lumbar vertebrae\n- **Format**: NIfTI (.nii.gz)\n\n### Dataset Structure\n\nEach sample in the JSONL file contains:\n```json\n{\n  \"image\": \"path/to/image.nii.gz\",\n  \"mask\": \"path/to/mask.nii.gz\",\n  \"label\": [\"organ1\", \"organ2\", ...],\n  \"modality\": \"CT\",\n  \"dataset\": \"CTPelvic1K\",\n  \"official_split\": \"train\",\n  \"patient_id\": \"patient_id\"\n}\n```\n\n## Usage\n\n### Load Metadata\n\n```python\nfrom datasets import load_dataset\n\n# Load the dataset\nds = load_dataset(\"MedOtter/CTPelvic1K\")\n\n# Access a sample\nsample = ds['train'][0]\nprint(f\"Patient ID: {sample['patient_id']}\")\nprint(f\"Image: {sample['image']}\")\nprint(f\"Mask: {sample['mask']}\")\nprint(f\"Labels: {sample['label']}\")\n```\n\n### Load Images\n\n```python\nfrom huggingface_hub import snapshot_download\nimport nibabel as nib\nimport os\n\n# Download the full dataset\nlocal_path = snapshot_download(\n    repo_id=\"MedOtter/CTPelvic1K\",\n    repo_type=\"dataset\"\n)\n\n# Load a sample\nsample = ds['train'][0]\nimage = nib.load(os.path.join(local_path, sample['image']))\nmask = nib.load(os.path.join(local_path, sample['mask']))\n\n# Get numpy arrays\nimage_data = image.get_fdata()\nmask_data = mask.get_fdata()\n\nprint(f\"Image shape: {image_data.shape}\")\nprint(f\"Mask shape: {mask_data.shape}\")\n```\n\n## Citation\n\n```bibtex\n@article{ctpelvic1k,\n  title={CTPelvic1K: A Large-Scale Dataset for Pelvic Bone Segmentation},\n  year={2023}\n}\n```\n\n## Dataset Homepage\n\nhttps://zenodo.org/record/4588403\n\n<!-- licence section rendered from the MedOtter card, 2026-09-06 -->\n## License\n\n**CC BY 4.0 (Zenodo 4588403: annotations, the CLINIC and CLINIC-metal images, models); the mirror carries only dataset6 CLINIC and dataset7 CLINIC-metal**\n\nRedistribution and commercial use are permitted under the terms below.\n\nSource of the terms: https://zenodo.org/records/4588403\n\nAudit note (verbatim from the MedOtter dataset card):\n\n> 2026-09-06 adjudication (settles the 2026-08-17 conditional): Zenodo record 4588403 (licence cc-by-4.0, access_right open) hosts 'Annotations and new collected clinical data in CTPelvic1K dataset': for datasets 1-5 only the annotations (*_mask_mappingback.tar.gz), for dataset6 (CLINIC) and dataset7 (CLINIC-metal) the images too. The mirror carries exactly dataset6 and dataset7 with their masks (103 + 14 cases, verified 2026-09-06) and nothing from datasets 1-5, whose images live under their own upstream licences (KiTS19 CC BY-NC-SA 4.0, BTCV Abdomen/Cervix no licence text, COLONOG CC BY 3.0, MSD CC BY-SA 4.0). Every byte the mirror hosts is therefore within the Zenodo CC BY 4.0 grant: redistribution yes, commercial use yes, attribution required. Adding any dataset1-5 images to the mirror re-opens this verdict (the BTCV subsets would make it no). Supersedes the 2026-08-17 conditional, which judged the whole seven-source dataset.\n\nPlease cite:\n\n```\nPengbo Liu, Hu Han, Yuanqi Du, Heqin Zhu, Yinhao Li, Feng Gu, Honghu Xiao, Jun Li, Chunpeng Zhao, Li Xiao, Xinbao Wu, S. Kevin Zhou. \"Deep Learning to Segment Pelvic Bones: Large-scale CT Datasets and Baseline Models.\" International Journal of Computer Assisted Radiology and Surgery (2021). https://doi.org/10.1007/s11548-021-02363-8\n```\n","encoding":"utf8","key":"README.md","name":"README.md","size":3631,"truncated":false}],"sourceObjectKey":"train.jsonl","starterCell":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/CTPelvic1K\")\nprint(ds)","starterSamplePath":"CTPelvic1K_dataset6_Anonymized_mask/dataset6_CLINIC_0001_mask_4label.nii.gz","totalObjects":237},"summary":"Pelvic bone CT with dense segmentations of the sacrum, left hip, right hip and lumbar vertebrae, from the CTPelvic1K dataset.","title":"CTPelvic1K CLINIC and CLINIC-metal: pelvic CT with sacrum, hip and lumbar vertebra segmentations (117 cases)","updatedAt":"2026-11-17T23:06:14.870Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/MedOtter/CTPelvic1K"}}