MCF-Net: Motion-Conditioned Multi-View Fusion for Myocardial Infarction Localization

Forum topic · 小凯 · 2026-07-18

Summary

MCF-Net is a novel motion-guided multi-view fusion framework for localizing myocardial infarction (MI) from echocardiography (Echo), proposed by Guang Yang, Wentian Xu, Siyu Wang, Betty Raman, Lei Li, and Vicente Grau (arXiv:2607.15268). MI remains a leading cause of death worldwide, and regional wall motion abnormality in Echo is a key diagnostic indicator. Prior learning-based myocardial motion analysis relies on handcrafted descriptors or densely supervised estimation, requiring extensive annotation. While Echo foundation models have improved vision-based analysis, most operate on single views and struggle with segment-level localization under view-dependent ambiguity, especially in apical views. MCF-Net extracts visual features using EchoPrime, a pretrained Echo foundation model shared across dual views, and models cardiac motion with extremely sparse supervision: a single annotated template frame is propagated across videos to initialize point tracking, avoiding dense labeling. Motion-derived segment-aware soft masks provide coarse spatial priors that selectively enhance challenging myocardial segments, and a motion-conditioned fusion mechanism integrates motion and vision across views without overriding strong appearance cues. MCF-Net achieves 72.4% F1 and 84.9% accuracy on segment-level MI localization, outperforming state-of-the-art motion-only, vision-only, and fusion baselines.

Paper Overview

Field: Computer Vision (Medical Imaging) Authors: Guang Yang, Wentian Xu, Siyu Wang, Betty Raman, Lei Li, Vicente Grau Published: 2026-07-16 arXiv: 2607.15268

Background

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.

Method: MCF-Net

To address these limitations, the authors propose MCF-Net, a motion-guided multi-view fusion framework that fuses myocardial motion cues with foundation model representations to localize infarction. Key components:

Vision features are extracted using EchoPrime, a pretrained Echo foundation model whose representations are shared across dual views.
Cardiac motion is modeled with extremely sparse supervision: a single annotated template frame is propagated across videos to initialize point tracking, avoiding dense annotation.
Motion-derived segment-aware soft masks provide coarse spatial priors that selectively enhance features of challenging myocardial segments.
A motion-conditioned fusion mechanism integrates motion and vision across views, refining predictions without overriding strong appearance cues.

Results

On segment-level MI localization, MCF-Net achieves:

72.4% F1 score
84.9% accuracy

These results outperform state-of-the-art motion-only, vision-only, and fusion baselines.

---

*Source: arXiv:2607.15268*