Enabling Ultra-Fast Cardiovascular Imaging Across Heterogeneous Clinical Environments with a Generalist Foundation Model and Multimodal Database

2025-12-25

AI Summary

Clinical cardiac magnetic resonance (CMR) imaging faces limitations in clinical adoption due to prolonged scan times and substantial heterogeneity across scanner vendors, acquisition protocols, and patient populations. To address this, we introduce MMCMR-427K—the first large-scale, multicenter, multimodal k-space database comprising 427K samples from 12 sites across five countries, six vendors, and nine pulse sequences. Building upon it, we propose CardioMM, a foundational reconstruction model for cardiovascular imaging. CardioMM integrates physics-driven differentiable reconstruction, joint semantic–physical modeling, and metadata-guided adaptive inference. It achieves, for the first time, vendor-, protocol-, and population-agnostic robust reconstruction without target-domain fine-tuning—demonstrating zero-shot generalization under 24× acceleration. It preserves diagnostic image quality and myocardial quantitative accuracy (e.g., LVEF error <1.2%), attaining state-of-the-art performance internally and stable generalization to unseen external sites. CardioMM significantly enhances CMR throughput and provides a unified reconstruction backbone for clinical deployment of multimodal CMR.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Multimodal cardiovascular magnetic resonance (CMR) imaging provides comprehensive and non-invasive insights into cardiovascular disease (CVD) diagnosis and underlying mechanisms. Despite decades of advancements, its widespread clinical adoption remains constrained by prolonged scan times and heterogeneity across medical environments. This underscores the urgent need for a generalist reconstruction foundation model for ultra-fast CMR imaging, one capable of adapting across diverse imaging scenarios and serving as the essential substrate for all downstream analyses. To enable this goal, we curate MMCMR-427K, the largest and most comprehensive multimodal CMR k-space database to date, comprising 427,465 multi-coil k-space data paired with structured metadata across 13 international centers, 12 CMR modalities, 15 scanners, and 17 CVD categories in populations across three continents. Building on this unprecedented resource, we introduce CardioMM, a generalist reconstruction foundation model capable of dynamically adapting to heterogeneous fast CMR imaging scenarios. CardioMM unifies semantic contextual understanding with physics-informed data consistency to deliver robust reconstructions across varied scanners, protocols, and patient presentations. Comprehensive evaluations demonstrate that CardioMM achieves state-of-the-art performance in the internal centers and exhibits strong zero-shot generalization to unseen external settings. Even at imaging acceleration up to 24x, CardioMM reliably preserves key cardiac phenotypes, quantitative myocardial biomarkers, and diagnostic image quality, enabling a substantial increase in CMR examination throughput without compromising clinical integrity. Together, our open-access MMCMR-427K database and CardioMM framework establish a scalable pathway toward high-throughput, high-quality, and clinically accessible cardiovascular imaging.

Problem

• Accelerates cardiovascular MRI scans while maintaining diagnostic quality across diverse clinical settings
• Addresses prolonged scan times and scanner/protocol heterogeneity limiting widespread clinical adoption
• Enables robust image reconstruction across varied equipment, protocols, and patient conditions

Innovation

• Generalist foundation model adapts to diverse cardiovascular imaging scenarios
• Multimodal k-space database enables robust reconstruction across heterogeneous environments
• Physics-informed model preserves diagnostic quality at ultra-fast 24x acceleration