Scholarlink

<p>Magnetocardiography (MCG) provides high-resolution spatiotemporal insights into cardiac electrophysiology but remains underutilized for&#xa0;left ventricular hypertrophy (LVH)&#xa0;diagnosis due to a lack of interpretable analytical tools. We propose a novel interpretable machine learning framework that systematically decodes MCG signals across four complementary domains: temporal waves, spatial waves, current source imaging, and dynamic characterization. To address class imbalance, we integrated&#xa0;Focal Loss&#xa0;into the XGBoost objective function. In a dataset of 481 subjects, our model achieved an&#xa0;AUC of 0.902&#xa0;in cross-validation and&#xa0;0.837&#xa0;in independent validation, significantly outperforming conventional baselines. Notably,&#xa0;SHapley Additive exPlanations (SHAP)&#xa0;identified T-wave magnetic polarity as the most influential predictor, offering new perspectives on the electrophysiological remodeling of hypertrophied myocardium. This framework bridges the gap between raw sensor data and clinical decision-making, providing a robust tool for automated LVH detection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning-enhanced MCG for LVH detection: a multi-domain feature selection approach

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月31日

查看全文
  • Xiaoxia Chen,
  • Xuanhao Xu,
  • Hong Shen,
  • Jinyang Wang,
  • Youhao Wang,
  • Lei Xia,
  • Jiaqi Wang,
  • Meifang Gao,
  • Chengxing Shen,
  • Yukun Luo

摘要

Magnetocardiography (MCG) provides high-resolution spatiotemporal insights into cardiac electrophysiology but remains underutilized for left ventricular hypertrophy (LVH) diagnosis due to a lack of interpretable analytical tools. We propose a novel interpretable machine learning framework that systematically decodes MCG signals across four complementary domains: temporal waves, spatial waves, current source imaging, and dynamic characterization. To address class imbalance, we integrated Focal Loss into the XGBoost objective function. In a dataset of 481 subjects, our model achieved an AUC of 0.902 in cross-validation and 0.837 in independent validation, significantly outperforming conventional baselines. Notably, SHapley Additive exPlanations (SHAP) identified T-wave magnetic polarity as the most influential predictor, offering new perspectives on the electrophysiological remodeling of hypertrophied myocardium. This framework bridges the gap between raw sensor data and clinical decision-making, providing a robust tool for automated LVH detection.