HVSMR-2.0: 3D cardiovascular MRI with whole-heart segmentations in congenital heart disease (60 patients)

Provider
Pace et al. (Boston Children's Hospital); mirrored by MedOtter
License
CC-BY-4.0 (source)
Size
546 files
Format
NIfTI, Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/MedOtter/HVSMR-2.0)
Access
Free and public. No account with the data provider is needed.
DOI
10.6084/m9.figshare.c.7074755.v2
Category
Health & Biomedical
Source
https://doi.org/10.6084/m9.figshare.c.7074755.v2

About this dataset

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).

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("MedOtter/HVSMR-2.0")
print(ds)

Data structure

File types in sampled files: NIfTI, Parquet

Files

Example file: data/train-00000-of-00001.parquet

Analyze this dataset without downloading it

On Scigantic, HVSMR-2.0: 3D cardiovascular MRI with whole-heart segmentations in congenital heart disease (60 patients) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Health & Biomedical datasets