发表机构
Saveetha School of Engineering; Government Kilpauk Medical College and Hospital（萨维塔工程学院; 政府基尔帕乌克医学院及医院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出 CardioFusion-AI 框架，通过信号处理前端与自适应融合策略，在多模态生理信号退化场景下实现鲁棒监测，验证了其在不同退化条件下的性能表现。
AI 中文摘要
可穿戴心电图（ECG）与光电容积脉搏波（PPG）传感器互为补充但各自脆弱：运动伪影、接触不良及传感器 dropout（信号丢失）会使其中一种或两种信号退化。假设两种模态同等可信的融合策略，在信号退化时可能比单一干净模态可靠性更低。本文提出 CardioFusion-AI 框架，其信号处理前端包括 R 峰与收缩期峰检测、Orphanidou 型信号质量指数、逐搏脉搏传导时间估计，已在 53 份真实重症监护记录（848 个窗口；ECG 的心率平均绝对误差为 1.61 bpm，PPG 为 2.78 bpm）及真实标注胎儿 ECG 数据库（R 峰 F1 值为 0.89-0.98）上验证。随后开展受控合成退化研究，在涵盖分级损坏与完全模态丢失的 6 种退化场景下，对比 8 种 ECG-PPG 融合策略，采用 5 个独立训练种子。注意力融合取得最低的描述性总误差（1.66±0.43 bpm）；两种自适应门控在完全模态丢失时均将权重重新分配至健康模态，但在分级退化下门控权重与信号质量的相关性极低（r=0.10-0.24）；信号质量调节在 PPG 缺失场景下取得特定改进（1.56±0.59 bpm），接近 1.48 bpm 的单模态上限。仅使用 5 个训练种子时，无成对比较通过 Holm 校正的显著性检验，因此报告效应量与置信区间。这些结果表明，模态可用性与模态质量是自适应融合的功能上不同的问题。
英文摘要
Wearable electrocardiogram (ECG) and photoplethysmogram (PPG) sensors are complementary but individually fragile: motion artifact, poor contact, and sensor dropout can degrade one or both signals. Fusion strategies that assume both modalities are equally trustworthy can become less reliable than a single clean modality under degradation. We present CardioFusion-AI, a framework whose signal-processing front end, including R-peak and systolic-peak detection, an Orphanidou-type signal-quality index, and beat-by-beat pulse transit time estimation, is validated on 53 real intensive-care recordings (848 windows; heart-rate mean absolute error 1.61 bpm for ECG and 2.78 bpm for PPG) and a real annotated fetal ECG database (R-peak F1 0.89-0.98). We then conduct a controlled synthetic degradation study comparing eight ECG-PPG fusion strategies across six degradation regimes spanning graded corruption and complete modality loss, using five independent training seeds. Attention fusion achieved the lowest descriptive overall error (1.66+/-0.43 bpm). Both adaptive gates reallocated weight toward the healthy modality under complete modality loss, but showed near-zero correlation between gate weight and signal quality under graded degradation (r = 0.10-0.24). Signal-quality conditioning produced a specific improvement under missing-PPG conditions (1.56+/-0.59 bpm), approaching the 1.48 bpm unimodal ceiling. With only five training seeds, no pairwise comparison survives Holm-corrected significance testing; effect sizes and confidence intervals are therefore reported. These results indicate that modality availability and modality quality are functionally distinct problems for adaptive fusion.
Comments7 pages, 4 figures. Under review at IEEE Journal of Biomedical and Health Informatics. Code: https://github.com/ka-cyber/CardioFusion-AI
