Scholarlink

<p>This study aims to develop and evaluate a wearable sensing and monitoring system for the diagnosis and therapy of obstructive sleep apnea (OSA), a disorder characterized by recurrent upper airway collapse during sleep that leads to sleep fragmentation and oxygen desaturation. The proposed system integrates multivariate biosensor modules—including flow, acceleration, temperature, humidity, and pH sensors—into an oral appliance designed to maintain the mandible in a forward position, thereby preventing airway collapse. A convolutional neural network was employed to analyze biosignals for detecting OSA-related events, bruxism, sleeping posture, and sleep patterns. In feasibility tests, the system accurately classified body postures, monitored heart rate variations, detected bruxism with up to 94% accuracy in non-bruxism segments and 89.47% in bruxism segments, and quantified sleep quality over extended monitoring periods. These results demonstrate that the proposed system offers a practical, non-invasive, and cost-effective alternative to conventional polysomnography or continuous positive airway pressure therapy, enabling continuous at-home screening and management of OSA, particularly in mild-to-moderate cases.</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

Sensing and monitoring system for diagnosis and therapy of obstructive sleep apnea

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月8日

查看全文
  • Won Ick Jang

摘要

This study aims to develop and evaluate a wearable sensing and monitoring system for the diagnosis and therapy of obstructive sleep apnea (OSA), a disorder characterized by recurrent upper airway collapse during sleep that leads to sleep fragmentation and oxygen desaturation. The proposed system integrates multivariate biosensor modules—including flow, acceleration, temperature, humidity, and pH sensors—into an oral appliance designed to maintain the mandible in a forward position, thereby preventing airway collapse. A convolutional neural network was employed to analyze biosignals for detecting OSA-related events, bruxism, sleeping posture, and sleep patterns. In feasibility tests, the system accurately classified body postures, monitored heart rate variations, detected bruxism with up to 94% accuracy in non-bruxism segments and 89.47% in bruxism segments, and quantified sleep quality over extended monitoring periods. These results demonstrate that the proposed system offers a practical, non-invasive, and cost-effective alternative to conventional polysomnography or continuous positive airway pressure therapy, enabling continuous at-home screening and management of OSA, particularly in mild-to-moderate cases.