"MMCMR-427K Part2"
Abstract
"Thank you for your interest in the MMCMR-427K database and CardioMM framework! Note: As the complete dataset exceeds 2 TB, it has been divided into two parts. This is Part 2, and Part 1 is available at: xxx.Motivation: Multimodal cardiovascular magnetic resonance (CMR) imaging offers comprehensive and non-invasive insights into cardiovascular disease (CVD) diagnosis and underlying mechanisms, but its widespread real-world adoption remains constrained by prolonged scan times and heterogeneity across medical environments. This underscores the urgent need to address a largely unexplored gap: a generalist reconstruction foundation model for ultra-fast CMR imaging\u2014one capable of adapting across diverse imaging scenarios and serving as the essential substrate for all downstream analyses.Database: To enable this goal, we curate MMCMR-427K, the largest and most comprehensive multimodal CMR k-space database to date. It comprises 427,465 multi-coil k-space data paired with structured metadata (approximately 3.5 TB), from 6,120 scans of 1,504 participants, spanning 13 international centers (four public repositories and nine clinical centers), 15 scanners (four vendors from low-field to ultra-high-field strengths), 12 CMR modalities, and 17 CVD categories across three populations (Asian, European, and North American).Engagement: Remarkably, in our organized CMRxRecon challenge series (2023\u20132025), more than 11,000 active participants from 125 countries and regions have engaged, reflecting the broad and growing global interest in next-generation CMR technologies.Official Website: https:\/\/github.com\/wangziblake\/CardioMM_MMCMR-427K"
Fields of Study & Subjects
Advanced MRI Techniques and Applications
Cardiac Imaging and Diagnostics
Medical Image Segmentation Techniques
Publication details
- Journal / source
- IEEE DataPort
- Publication date
- 15 March 2026
- Language
- EN
- Document type
- Dataset