期刊
IEEE ACCESS
卷 8, 期 -, 页码 46131-46140出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2020.2978391
关键词
Sleep-stage classification; multi-biosignal interface; rule-based decision tree; feature extraction stage; readout integrated circuit; wearable device
资金
- Institute of Information and Communications Technology Planning and Evaluation (IITP) - Ministry of Science and ICT, South Korea [2018-0-00756, 2019-0-00208]
- Samsung Electronics
- National Research Foundation of Korea [2017M1A2A2087833]
This paper presents a wearable multi-biosignal wireless interface for sleep analysis. It enables comfortable sleep monitoring with direct sleep-stage classification capability while conventional analytic interfaces including the Polysomnography (PSG) require complex post-processing analyses based on heavy raw data, need expert supervision for measurements, or do not provide comfortable fit for long-time wearing. The proposed multi-biosignal interface consists of electroencephalography (EEG), electromyography (EMG), and electrooculography (EOG). A readout integrated circuit (ROIC) is designed to collect three kinds of bio-potential signals through four internal readout channels, where their analog feature extraction circuits are included together to provide sleep-stage classification directly. The designed multi-biosignal sensing ROIC is fabricated in a 180-nm complementary metal-oxide-semiconductor (CMOS) process. For system-level verification, its low-power headband-style analytic device is implemented for wearable sleep monitoring, where the direct sleep-stage classification is performed based on a decision tree algorithm. It is functionally verified by comparison experiments with post-processing analysis results from the OpenBCI module, whose sleep-stage detection shows reasonable correlation of 74% for four sleep stages.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据