{"success":true,"data":{"author":"Noelia Vallez et al. (Hospital General de Ciudad Real); mirrored by MedOtter","bucketAccessType":"public","bucketName":"MedOtter/BUS-UCLM","bucketPrefix":"","category":"Health & Biomedical","cloudProvider":"huggingface","createdAt":"2026-11-17T19:29:33.469Z","description":"BUS-UCLM is a breast ultrasound lesion segmentation dataset collected at Hospital General de Ciudad Real, Spain, on a Siemens ACUSON S2000 with the 18L6 HD probe. It holds 683 images from 38 patients: 174 with benign lesions, 90 malignant and 419 normal, at roughly 856x606 pixels. The original masks are RGB color-coded by class (green benign, red malignant); this mirror stores one binary lesion mask per image and keeps the class in a class_label column, alongside patient_id and native resolution, in a Parquet file under data/. There is no official split; the authors' helper repository suggests a 33/5 patient partition. About 330 MB (measured 2026-11-17). Use for breast lesion segmentation and benign versus malignant classification in ultrasound. Mirrored on Hugging Face by MedOtter from the authors' release, with a card documenting provenance and any normalization. The authors' Mendeley Data record (doi 10.17632/7fvgj4jsp7) is CC BY 4.0; cite Vallez et al., Scientific Data 2025.","displayTitle":"BUS-UCLM: 683 breast ultrasound images from 38 patients with lesion masks (benign, malignant, normal)","doi":"10.17632/7fvgj4jsp7.3","huggingFaceUrl":"https://huggingface.co/datasets/MedOtter/BUS-UCLM","id":"78e747b5-fb5b-5440-a9df-d83433562165","license":"CC-BY-4.0","licenseEvidenceUrl":"https://data.mendeley.com/datasets/7fvgj4jsp7/3","objectCountEstimate":3,"probeStatus":"ok","referenceUrl":"https://doi.org/10.17632/7fvgj4jsp7.3","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/BUS-UCLM\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"MedOtter/BUS-UCLM\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=MedOtter, license:cc-by-4.0, size_categories:n<1K, downloads=250"],"format":"mixed","generatedAt":"2026-11-17T19:40:30.918Z","layout":{"example":"data/train-00000-of-00001.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"data/train-00000-of-00001.parquet","truncatedAt":512}],"sampleObjects":["data/train-00000-of-00001.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n  - image-segmentation\ntags:\n  - medical\n  - ultrasound\n  - breast\n  - segmentation\n  - bus-uclm\nsize_categories:\n  - n<1K\n---\n\n# BUS-UCLM — Breast Ultrasound Lesion Segmentation Dataset\n\nRe-hosted mirror of the BUS-UCLM dataset (Vallez et al., *Scientific Data* 2025),\ncollected at Hospital General de Ciudad Real, Spain on a Siemens ACUSON S2000\nwith the 18L6 HD probe. Originally distributed via Mendeley Data\n(DOI: [10.17632/7fvgj4jsp7.3](https://doi.org/10.17632/7fvgj4jsp7.3)).\n\nThis mirror is intended for use with [EasyMedSeg](https://github.com/) and\nprovides:\n\n- A single canonical source (the original Mendeley S3 cache returns 403 to\n  direct fetches; we mirror the canonical `images/` + `masks/` + `INFO.csv`).\n- Parquet schema with one row per image and a single binary mask.\n- The original mask format is **RGB-color-coded by class** (green for benign,\n  red for malignant, all-zero for normal). This mirror **collapses each mask to\n  a binary `L`-mode image** (any non-zero pixel → foreground). The class label\n  is preserved in the `class_label` column.\n\n## Composition\n\n| Class       | Images |\n|-------------|-------:|\n| benign      | 174    |\n| malignant   | 90     |\n| normal      | 419    |\n| **Total**   | **683**|\n\n683 images from 38 patients (38 distinct patient prefixes in the filenames).\nStandard image dimensions ~856×606 px (varies by case; full per-image\nresolution is preserved in the `resolution` column).\n\n## Splits\n\nBUS-UCLM has **no official train/val/test split**. This release ships a single\n`train` split. The authors' [helper repo](https://github.com/noeliavallez/BUS-UCLM-Dataset)\nsuggests a 33/5 patient partition with five test prefixes (`COPE`, `ANFO`,\n`ELCO`, `CRCI`, `FLKA`); reproduce it downstream with a `patient_id`-based\nfilter if needed.\n\n## Schema\n\n| Column          | Type      | Description                                                |\n|-----------------|-----------|------------------------------------------------------------|\n| `image`         | `Image`   | Source PNG (RGB)                                           |\n| `mask`          | `Image`   | Binary mask (`L` mode, 0/255) — collapsed from RGB-coded source |\n| `image_id`      | `string`  | e.g., `\"ALWI_000\"`                                         |\n| `patient_id`    | `string`  | First underscore token (e.g., `\"ALWI\"`)                    |\n| `class_label`   | `string`  | `\"benign\"` / `\"malignant\"` / `\"normal\"`                    |\n| `resolution`    | `string`  | Native resolution string from `INFO.csv` (e.g., `\"856x606\"`) |\n| `has_doppler`   | `bool`    | Whether the image is a Doppler frame                       |\n| `has_marks`     | `bool`    | Whether the source image carries on-screen annotations (calipers, arrows) |\n| `has_combined`  | `bool`    | Whether the source image is a \"combined\" (montage) acquisition |\n\n## License\n\nCC BY 4.0, inherited from the upstream Mendeley release.\n\n## Citation\n\n```bibtex\n@article{vallez2025bus,\n  title   = {BUS-UCLM: Breast ultrasound lesion segmentation dataset},\n  author  = {Vallez, Noelia and Bueno, Gloria and Deniz, Oscar and Rienda, Miguel Angel and Pastor, Carlos},\n  journal = {Scientific Data},\n  volume  = {12},\n  number  = {1},\n  pages   = {242},\n  year    = {2025},\n  doi     = {10.1038/s41597-025-04562-3}\n}\n```\n","encoding":"utf8","key":"README.md","name":"README.md","size":3348,"truncated":false}],"sourceObjectKey":"data/train-00000-of-00001.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/BUS-UCLM\")\nprint(ds)","starterSamplePath":"data/train-00000-of-00001.parquet","totalObjects":3},"summary":"BUS-UCLM is a breast ultrasound lesion segmentation dataset collected at Hospital General de Ciudad Real, Spain, on a Siemens ACUSON S2000 with the 18L6 HD probe.","title":"BUS-UCLM: 683 breast ultrasound images from 38 patients with lesion masks (benign, malignant, normal)","updatedAt":"2026-11-17T23:15:02.913Z","visibility":"public","websiteUrl":"https://github.com/noeliavallez/BUS-UCLM-Dataset"}}