Article record

长程相互依赖感知的信号分析用于快速准确预测阻塞性睡眠呼吸暂停

Long-range interdependencies aware signal analysis for rapid and accurate obstructive sleep apnea prediction.

AI/ML咽喉科IF 18.9Q1

文献信息

PMID	42907459
原文	在 PubMed 查看原文 ↗
发表日期	2026
作者	Zhenkun Wang 查看其他 20 位作者 Chenzhong Yin；Anzhe Cheng；Xinghe Chen；Mingxi Cheng；Li An；Haoqing Wang；Haofan Sun；Li Ding；Stefan Mihaicuta；Lucretia Udrescu；David Gozal；Félix Del Campo；Fernando Vaquerizo-Villar；Roberto Hornero；Jan Hedner；Winfried Randerath；David M Mannino；Shahin Nazarian；Mihai Udrescu；Paul Bogdan
作者单位	Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, USA. 查看其他 19 条单位信息 Corwin D. Denney Research Center, Department of Biomedical Engineering, Viterbi School of Engineering, University of Southern California, Los Angeles, CA, USA.；School for Engineering of Matter, Transport and Energy, Arizona State University, Tempe, AZ, USA.；Department of Pulmonology, "Victor Babeş" University of Medicine and Pharmacy Timişoara, Timişoara, Romania.；Center for Research and Innovation in Personalized Medicine of Respiratory Diseases (CRIPMRD), "Victor Babeş" University of Medicine and Pharmacy Timişoara, Timişoara, Romania.；Department I - Clinical Pharmacy and Drug Analysis, "Victor Babeş" University of Medicine and Pharmacy Timişoara, Timişoara, Romania.；Center for Drug Data Analysis, Cheminformatics, and the Internet of Medical Things (CDDA-CIMT), "Victor Babeş" University of Medicine and Pharmacy Timişoara, Timişoara, Romania.；Departments of Pediatrics, Biomedical Sciences and Dentistry and Maxillofacial Surgery, Joan C. Edwards School of Medicine, Marshall University, Huntington, WV, USA.；Biomedical Engineering Group, University of Valladolid, Valladolid, Spain.；CIBER de Bioingeniería, Biomateriales y Nanomedicina, Instituto de Salud Carlos III, Valladolid, Spain.；Hospital Universitario Río Hortega. Valladolid, Valladolid, Spain.；Pulmonary Department, Sleep Disorders Center, Sahlgrenska University Hospital, Gothenburg, Sweden.；Center of Sleep and Wake Disorders, Sahlgrenska Academy, Gothenburg University, Göteborg, Sweden.；Institute Für Pneumologie an der Universität Zu Köln, Cologne, Germany.；College of Medicine, University of Kentucky, Lexington, KY, USA.；COPD Foundation, Miami, FL, USA.；Department of Computer and Information Technology, Politehnica University Timişoara, Timişoara, Romania. mihai.udrescu@cs.upt.ro.；Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, USA. pbogdan@usc.edu.；Center for Complex Particle Systems (COMPASS), University of Southern California, Los Angeles, CA, USA. pbogdan@usc.edu.；Thomas Lord Department of Computer Science, University of Southern California, Los Angeles, CA, USA. pbogdan@usc.edu.
期刊	Nature communications
SCI 分区	Q1
IF	18.9
研究类型	AI/ML · 临床
所属专科	咽喉科

中文摘要

阻塞性睡眠呼吸暂停以呼吸频繁中断和间歇性低氧为特征，因其可能增加多器官系统发病率和死亡率而成为公共卫生问题。尽管多导睡眠图是诊断的金标准，但其操作繁琐、成本高昂且复杂。在此，我们提出一个生物几何感知的人工智能框架，处理生理信号以预测睡眠呼吸暂停严重程度。该框架提取多重分形耦合特征，捕捉生理过程之间的长程依赖关系，并将其整合到深度学习中进行分类。我们在五个异质性队列中验证了该方法，共纳入超过35,000名受试者。该方法实现了稳健且一致的性能，曲线下面积值超过0.91，较最先进模型提升8-18%。提取的耦合表征表现出快速收敛，并在异质性队列中保持一致的性能。该框架提供了生理动力学的稳健表征，并支持睡眠呼吸障碍自动化分析的未来发展。

英文摘要

Obstructive sleep apnea, characterized by frequent interruptions in breathing and intermittent hypoxia, is a public health concern due to its potential to increase multiple organ system morbidity and mortality. Although polysomnography is the gold standard for diagnosis, it is laborious, costly, and complex. Here, we present a biological geometry-aware artificial intelligence framework that processes physiological signals to predict sleep apnea severity. This framework extracts multifractal coupling features that capture long-range dependencies among physiological processes and integrates them into deep learning for classification. We validate this approach across five heterogeneous cohorts comprising more than 35,000 subjects. The method achieves robust and consistent performance, with area-under-the-curve values exceeding 0.91 and improvements of 8-18% over state-of-the-art models. The extracted coupling representations exhibit rapid convergence and consistent performance across heterogeneous cohorts. This framework provides a robust representation of physiological dynamics and supports future developments in automated analysis of sleep-disordered breathing.