Motion-Conditioned Multi-View Fusion for Myocardial Infarction Localization from Echocardiography

2026-07-16

AI Summary

This study addresses the view-dependent ambiguity of single-view echocardiography in myocardial infarction segment localization and the limited clinical applicability of existing methods that rely on densely annotated data. To overcome these challenges, the authors propose MCF-Net, a novel framework that, for the first time, integrates myocardial motion modeling under sparse supervision with visual representations from EchoPrime—a pretrained foundation model for echocardiography. The approach employs motion-guided soft masking to enhance features of critical segments and introduces a motion-conditioned multi-view fusion mechanism to effectively aggregate cross-view information. Notably, MCF-Net operates without dense annotations and achieves state-of-the-art performance on segment-level localization, attaining an F1 score of 72.4% and an accuracy of 84.9%, significantly outperforming current motion-based, vision-based, and fusion-based baselines.

Abstract

Myocardial infarction (MI) remains a leading cause of mortality worldwide. Echocardiography (Echo) is a widely available modality for MI assessment, where regional wall motion abnormality is a key indicator. Prior learning based methods for myocardial motion analysis often use handcrafted descriptors or densely supervised estimation, but the need for extensive annotation limits applicability. Foundation models have recently improved vision-based Echo analysis; however, most methods operate on single views and segment-level localization remains unreliable under view-dependent ambiguity, especially in apical views. To address this, we propose MCF-Net, a novel motion-guided multi-view fusion framework that fuses myocardial motion cues with foundation model representations to localize infarction. Visual features are extracted using EchoPrime, a pretrained Echo foundation model shared across dual views. Cardiac motion is modeled with extremely sparse supervision: a single annotated template frame is transferred across videos to initialize point tracking, avoiding dense labels. Motion-derived segment-aware soft masks provide coarse spatial priors that selectively enhance features for challenging myocardial segments. A motion-conditioned fusion mechanism then integrates motion and vision across views, refining predictions without overriding strong appearance cues. On segment-level MI localization, MCF-Net achieves 72.4\% F1 and 84.9\% accuracy, outperforming state-of-the-art motion-only, vision-only, and fusion baselines.

Problem

• myocardial infarction
• echocardiography
• multi-view fusion
• wall motion abnormality
• segment-level localization

Innovation

• motion-conditioned fusion
• multi-view echocardiography
• foundation model
• sparse supervision
• myocardial infarction localization