PhysCL: Knowledge-Aware Contrastive Learning of Physiological Signal Models for Cuff-Less Blood Pressure Estimation

doi:10.1109/JBHI.2025.3554495

	PhysCL: Knowledge-Aware Contrastive Learning of Physiological Signal Models for Cuff-Less Blood Pressure Estimation
	Liu, Renju 1,2; Shen, Jianfei 1; Gu, Yang 1; Chen, Yiqiang 1; Zhang, Jiling 3; Wu, Qingyu 1; Xu, Chenyang 4; Fan, Feiyi 1
	2025-07-01
发表期刊	IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
ISSN	2168-2194
卷号	29 期号:7 页码:4728-4740
摘要	Training deep learning models for photoplethysmography(PPG)-based cuff-less blood pressure estimation often requires a substantial amount of labeled data collected through sophisticated medical instruments, posing significant challenges in practical applications. To address this issue, we propose Physiological Knowledge-Aware Contrastive Learning (PhysCL), a novel approach designed to reduce the dependence on labeled PPG data while improving blood pressure estimation accuracy. Specifically, PhysCL tackles the semantic consistency problem in contrastive learning by introducing a knowledge-aware augmentation bank, which generates positive physiological signal pairs using knowledge-based constraints during the contrastive pair generation. Additionally, we propose a contrastive feature reconstruction method to enhance feature diversity and prevent model collapse through feature re-sampling and re-weighting. We evaluate PhysCL on data from 106 subjects across the MIMIC III, MIMIC IV, and UQVS datasets under cross-dataset validation settings, comparing it against state-of-the-art contrastive learning methods and blood pressure estimation models. PhysCL achieves an average mean absolute error of 9.5/5.9 mmHg (systolic/diastolic) across the three datasets, using only 2% labeled data combined with 98% unlabeled data for pre-training and 5 samples for personalization, which represents a 6.2% /4.3% improvement, respectively, over the current best supervised methods. The ablation study provides further convincing evidence that the unlabeled data can be utilized to improve the existing cuff-less blood pressure estimation models and shed light on unsupervised contrastive learning for physiological signals.
关键词	Blood pressure Physiology Contrastive learning Data augmentation Semantics Feature extraction Estimation Time series analysis Training Data models Blood pressure estimation physiological signal processing representation learning self-supervised learning
DOI	10.1109/JBHI.2025.3554495
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62101530] ; Beijing Municipal Science & Technology Commission[Z221100002722009] ; Youth Innovation Promotion Association CAS[2021101] ; AI Dream Project[SZSM202401] ; AI Dream Project[SZSM202403]
WOS研究方向	Computer Science ; Mathematical & Computational Biology ; Medical Informatics
WOS类目	Computer Science, Information Systems ; Computer Science, Interdisciplinary Applications ; Mathematical & Computational Biology ; Medical Informatics
WOS记录号	WOS:001523482700009
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/42021
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Fan, Feiyi
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China 2.Beijing Normal Univ Hong Kong Baptist Univ United, Dept Comp Sci, Zhuhai 519087, Peoples R China 3.Zhejiang Univ, Sch Software Technol, Hangzhou 310058, Peoples R China 4.Tianjin Univ, Sch Elect & Informat Engn, Tianjin 300072, Peoples R China
推荐引用方式 GB/T 7714	Liu, Renju,Shen, Jianfei,Gu, Yang,et al. PhysCL: Knowledge-Aware Contrastive Learning of Physiological Signal Models for Cuff-Less Blood Pressure Estimation[J]. IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS,2025,29(7):4728-4740.
APA	Liu, Renju.,Shen, Jianfei.,Gu, Yang.,Chen, Yiqiang.,Zhang, Jiling.,...&Fan, Feiyi.(2025).PhysCL: Knowledge-Aware Contrastive Learning of Physiological Signal Models for Cuff-Less Blood Pressure Estimation.IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS,29(7),4728-4740.
MLA	Liu, Renju,et al."PhysCL: Knowledge-Aware Contrastive Learning of Physiological Signal Models for Cuff-Less Blood Pressure Estimation".IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS 29.7(2025):4728-4740.