A Multimodal and Explainable Machine Learning Approach to Diagnosing Multi-Class Ejection Fraction from Electrocardiograms
A Multimodal and Explainable Machine Learning Approach to Diagnosing Multi-Class Ejection Fraction from Electrocardiograms
一种用于诊断多分类射血分数的模态融合与可解释机器学习方法
Abstract: Left ventricular ejection fraction (LVEF) assessment depends on echocardiography, limiting access in primary care and resource-constrained settings. We developed a multimodal machine-learning framework that combines engineered 12-lead ECG timeseries features with structured EHR variables to classify LVEF into four clinically used strata: normal (>50%), mildly reduced (40-50%), moderately reduced (30-40%), and severely reduced (<30%).
摘要： 左心室射血分数（LVEF）的评估依赖于超声心动图，这限制了其在基层医疗和资源受限环境中的应用。我们开发了一种多模态机器学习框架，将工程化的12导联心电图（ECG）时间序列特征与结构化的电子健康记录（EHR）变量相结合，将LVEF分为四个临床常用等级：正常（>50%）、轻度降低（40-50%）、中度降低（30-40%）和重度降低（<30%）。
To support model explainability, we identified the most influential ECG and EHR features via SHAP attributions. Using retrospective data from Hartford HealthCare, we trained XGBoost models on 36,784 ECG-echocardiogram pairs from 30,952 outpatients and evaluated temporal generalizability on 19,966 ECGs from a subsequent period.
为了支持模型的可解释性，我们通过SHAP归因分析识别出了最具影响力的心电图和电子健康记录特征。利用哈特福德医疗保健系统（Hartford HealthCare）的回顾性数据，我们对来自30,952名门诊患者的36,784对心电图-超声心动图数据进行了XGBoost模型训练，并评估了其在随后一段时间内19,966份心电图数据上的时间泛化能力。
The multimodal model achieved one-vs-rest AUROCs of 0.95 (severe), 0.92 (moderate), 0.82 (mild), and 0.91 (normal), outperforming ECG-only and EHR-only baselines, and maintained performance under temporal validation. This work supports ECG-based, multimodal LVEF stratification as a practical screening and triage aid to prioritize confirmatory imaging where resources are limited.
该多模态模型在“一对多”（one-vs-rest）分类任务中取得了0.95（重度）、0.92（中度）、0.82（轻度）和0.91（正常）的AUROC值，表现优于仅使用心电图或仅使用电子健康记录的基准模型，并在时间验证中保持了良好的性能。这项研究表明，基于心电图的多模态LVEF分层可以作为一种实用的筛查和分诊辅助手段，在资源有限的情况下优先安排确诊影像检查。
Paper Details:
- Authors: Catherine Ning, Yu Ma, Cindy Beini Wang, Sean McMahon, Joseph Radojevic, Steven Zweibel, Dimitris Bertsimas
- arXiv ID: 2604.25942
- Subject: Machine Learning (cs.LG)
论文详情：
- 作者： Catherine Ning, Yu Ma, Cindy Beini Wang, Sean McMahon, Joseph Radojevic, Steven Zweibel, Dimitris Bertsimas
- arXiv编号： 2604.25942
- 学科： 机器学习 (cs.LG)