{"success":true,"data":{"author":"Pace et al. (Boston Children's Hospital); mirrored by MedOtter","bucketAccessType":"public","bucketName":"MedOtter/HVSMR-2.0","bucketPrefix":"","category":"Health & Biomedical","cloudProvider":"huggingface","createdAt":"2026-11-17T19:29:33.270Z","description":"HVSMR-2.0 is a 3D cardiovascular magnetic resonance (CMR) dataset for whole-heart segmentation in congenital heart disease. It contains 60 scans from 60 patients aged under 1 to 52 years (mean 11.6) acquired at Boston Children's Hospital on a 1.5 T Philips Achieva: axial steady-state free precession (SSFP), free-breathing with a respiratory navigator and prospective ECG gating, static rather than cine, with mixed gadolinium contrast and near-isotropic voxels of about 0.73x0.73x0.81 mm. Labels mark eight disjoint structures: left and right ventricles, left and right atria, aorta, pulmonary artery, superior and inferior vena cava; there is no myocardium label. Cases with a common atrium or a single ventricle have the absent chamber left empty by design. The repository keeps the three figshare versions (orig/, cropped/, cropped_norm/) as NIfTI, plus hvsmr_clinical.csv (per-patient diagnosis flags such as common atrium and single ventricle) and hvsmr_technical.csv; about 3.5 GB (measured 2026-11-17). 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' figshare items are CC BY 4.0; cite Pace et al., Scientific Data 11, 721 (2024).","displayTitle":"HVSMR-2.0: 3D cardiovascular MRI with whole-heart segmentations in congenital heart disease (60 patients)","doi":"10.6084/m9.figshare.c.7074755.v2","huggingFaceUrl":"https://huggingface.co/datasets/MedOtter/HVSMR-2.0","id":"84ba3b48-5c17-5a14-b44d-8c2545ad4744","license":"CC-BY-4.0","licenseEvidenceUrl":"https://api.figshare.com/v2/articles/25226360","objectCountEstimate":546,"probeStatus":"ok","referenceUrl":"https://doi.org/10.6084/m9.figshare.c.7074755.v2","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/HVSMR-2.0\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"MedOtter/HVSMR-2.0\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=MedOtter, license:cc-by-4.0, size_categories:n<1K, format:parquet, downloads=1145"],"format":"mixed","generatedAt":"2026-11-17T19:40:23.771Z","layout":{"example":"data/train-00000-of-00001.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"50415231150415a40815fc034c1578150012000092041c04000000706174300d0800310d0800320d0800330d0800340d0800350d0800360d0800370d0800380d08043905095004313011090031110900321109003311090034110900351109003611090037110900381109115a0032115a0032115a0032115a0032115a0032115a0032115a0032115a0032115a0032115a0032115a0033115a0033115a0033115a0033115a0033115a0033115a0433360da20033115a0033115a0033115a0034115a0034115a0034115a0034115a0034115a0034115a0034115a0034115a0034115a0034115a0035115a0035115a0035115a0035115a0035115a0035115a0035115a0035115a2835380500000070617435391500157015742c1578151015061506000038dc020000007801061140200c44611c48a22c4ce33c50244d54655d58a66d5ce77d60288e64699e68aaae6cebbe702ccf746ddf78aeef000000150415e00315e8034c15781500120000f001f0ef000000000100000002000000030000000400000005000000060000000700000008000000090000000a0000000b0000000c0000000d0000000e0000000f000000100000001100000012000000130000001400000015000000160000001700000018000000190000001a0000001b0000001c0000001d0000001e0000001f000000200000002100000022000000230000","encoding":"hex","key":"data/train-00000-of-00001.parquet","truncatedAt":512}],"sampleObjects":["cropped/pat0_cropped.nii.gz","cropped/pat0_cropped_seg.nii.gz","cropped/pat0_cropped_seg_endpoints.nii.gz","cropped/pat10_cropped.nii.gz","cropped/pat10_cropped_seg.nii.gz"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\npretty_name: HVSMR-2.0 — whole-heart CMR in congenital heart disease\ntags:\n- medical-imaging\n- mri\n- cardiac\n- congenital-heart-disease\n- whole-heart-segmentation\n- 3d\ntask_categories:\n- image-segmentation\nconfigs:\n- config_name: default\n  data_files:\n  - split: train\n    path: data/train-*\ndataset_info:\n  features:\n  - name: sample_id\n    dtype: string\n  - name: pat\n    dtype: int32\n  - name: split\n    dtype: string\n  - name: image\n    dtype: image\n  - name: mask\n    dtype: image\n  - name: overlay\n    dtype: image\n  - name: optional_zone\n    dtype: image\n  - name: present_classes\n    dtype: string\n  - name: absent_classes\n    dtype: string\n  - name: n_classes\n    dtype: int32\n  - name: hvsmr2016\n    dtype: string\n  - name: pace_media2022\n    dtype: string\n  - name: severity\n    dtype: string\n  - name: age_years\n    dtype: string\n  - name: diagnoses\n    dtype: string\n  - name: shape_orig\n    dtype: string\n  - name: shape_cropped\n    dtype: string\n  - name: spacing_mm\n    dtype: string\n  - name: num_axial_slices\n    dtype: int32\n  - name: slice_index\n    dtype: int32\n  - name: gt_voxels\n    dtype: int64\n  - name: optional_zone_voxels\n    dtype: int64\n  - name: te_ms\n    dtype: float32\n  - name: tr_ms\n    dtype: float32\n  - name: flip_angle_deg\n    dtype: float32\n  - name: bandwidth_hz\n    dtype: float32\n  splits:\n  - name: train\n    num_bytes: 7292850\n    num_examples: 60\n  download_size: 7303916\n  dataset_size: 7292850\n---\n\n# HVSMR-2.0\n\nRaw mirror of **HVSMR-2.0**, a 3D cardiovascular MR dataset for whole-heart\nsegmentation in **congenital heart disease**, from the authors' figshare\ncollection ([10.6084/m9.figshare.c.7074755.v2](https://doi.org/10.6084/m9.figshare.c.7074755.v2);\nPace et al., *Scientific Data* 11:721, 2024).\n\n**60 CMR scans from 60 patients** (ages <1–52 y, mean 11.6), acquired at Boston\nChildren's Hospital on a 1.5 T Philips Achieva: axial **SSFP**, free-breathing\nwith respiratory navigator, prospective ECG gating. This is a static whole-heart\nacquisition, **not cine**. Many but not all patients received IV gadolinium, so\ncontrast is mixed by design. Near-isotropic voxels (avg 0.73×0.73×0.81 mm).\nAll volumes are **PIR**-oriented.\n\n## Classes — 8 foreground structures, strictly disjoint\n\n| Label | Structure | Label | Structure |\n|---|---|---|---|\n| 0 | background | 5 | AO — aorta |\n| 1 | LV — left ventricle | 6 | PA — pulmonary artery |\n| 2 | RV — right ventricle | 7 | SVC — superior vena cava |\n| 3 | LA — left atrium | 8 | IVC — inferior vena cava |\n| 4 | RA — right atrium | | |\n\nOne integer per voxel — no nesting, no overlap. Verified across all 180\nmasks: the value union is exactly `{0..8}`.\n\n**There is no myocardium label.** HVSMR 2016 segmented blood pool + ventricular\nmyocardium; 2.0 deliberately dropped myocardium in favour of the 8 chambers and\ngreat vessels. Do not expect the two label schemes to be compatible.\n\n### 13 of 60 cases have empty classes — this is anatomy, not annotation error\n\n- **RA (4) absent in 10 cases**: pat 16, 20, 30, 39, 45, 47, 49, 50, 56, 58 —\n  these are *exactly* the ten cases flagged `CommonAtrium` in\n  `hvsmr_clinical.csv` (verified), where the common atrium is labelled LA.\n- **RV (2) absent in 6 cases**: pat 50, 52, 56, 57, 58, 59 — single-ventricle\n  cases, where the single ventricle is labelled LV. Five of the six are flagged\n  `SingleVentricle` in the clinical CSV; the sixth is pat59, the one case with\n  no clinical row at all.\n- Both absent in 3: pat 50, 56, 58. The remaining **47 cases carry all 8**.\n\nThe empty classes are therefore genuine anatomy, not missing annotation.\n\nMacro-averaged Dice must **skip classes absent from the ground truth** rather\nthan score them 0. `present_classes` / `absent_classes` in `train.jsonl` give\nthis per case.\n\n## Three variants — pick one, do not mix\n\n| Directory | Image | Intensities | Size |\n|---|---|---|---|\n| `orig/` | full volume, cropped at the chin for de-identification | raw (~[0, 7500]) | 1.85 GB |\n| `cropped/` | tight crop around the heart, same resolution | raw | 0","encoding":"utf8","key":"README.md","name":"README.md","size":12106,"truncated":true}],"sourceObjectKey":"data/train-00000-of-00001.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"MedOtter/HVSMR-2.0\")\nprint(ds)","starterSamplePath":"cropped/pat0_cropped.nii.gz","totalObjects":546},"summary":"HVSMR-2.0 is a 3D cardiovascular magnetic resonance (CMR) dataset for whole-heart segmentation in congenital heart disease.","title":"HVSMR-2.0: 3D cardiovascular MRI with whole-heart segmentations in congenital heart disease (60 patients)","updatedAt":"2026-11-17T23:13:44.711Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/MedOtter/HVSMR-2.0"}}