[论文] Motion-Conditioned Multi-View Fusion for Myocardial Infarction Localiz...

论文概要

研究领域: CV 作者: Guang Yang, Wentian Xu, Siyu Wang, Betty Raman, Lei Li, Vicente Grau 发布时间: 2026-07-16 arXiv: 2607.15268

中文摘要

心肌梗死（MI）仍是全球主要死亡原因之一。超声心动图（Echo）是一种广泛可用的MI评估方式，其中区域壁运动异常是关键指标。先前基于学习的心肌运动分析方法通常使用手工描述符或密集监督估计，但广泛标注的需求限制了其适用性。基础模型最近改善了基于视觉的Echo分析；然而，大多数方法在单视角上操作，且在视角相关歧义下，节段级定位仍然不可靠，尤其是在心尖视角。为解决这一问题，我们提出MCF-Net，一种新颖的运动引导多视角融合框架，将心肌运动线索与基础模型表征融合以定位梗死。视觉特征使用EchoPrime提取，这是一个跨双视角共享的预训练Echo基础模型。心脏运动以极稀疏监督建模：单个标注模板帧跨视频迁移以初始化点追踪，避免了密集标注。运动导出的节段感知软掩码提供粗略空间先验，选择性地增强具有挑战性心肌节段的特征。然后，一种运动条件融合机制跨视角整合运动和视觉，在不覆盖强外观线索的情况下精炼预测。在节段级MI定位上，MCF-Net达到72.4% F1和84.9%准确率，优于最先进的仅运动、仅视觉和融合基线。

原文摘要

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 infarcti...

--- *自动采集于 2026-07-18*

#论文 #arXiv #CV #小凯