여운홍 (Georgia Tech) | 한빛사논문 > 한빛사

한빛사논문

조회512

Georgia Tech

CV File 899 kb

Biosens. Bioelectron., Dec 28 2023, 248 S0956-5663(23)00925-9 | https://doi.org/10.1016/j.bios.2023.115983 Continuous real-time assessment of acute cognitive stress from cardiac mechanical signals captured by a skin-like patch

Nathan Zavanelli ^a,b, Sung Hoon Lee ^b,c, Matthew Guess ^a,b, Woon-Hong Yeo ^a,b,d,e

^aGeorge W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, 30024, USA

^bIEN Center for Wearable Intelligent Systems and Healthcare at the Institute for Electronics and Nanotechnology, Georgia Institute of Technology, Atlanta, GA, 30332, USA

^cSchool of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, 30332, USA

^dWallace H. Coulter Department of Biomedical Engineering, Georgia Tech and Emory University School of Medicine, Atlanta, GA, 30332, USA

^eParker H. Petit Institute for Bioengineering and Biosciences, Institute for Materials, Institute for Robotics and Intelligent Machines, Georgia Institute of Technology, Atlanta, GA, 30332, USA

Corresponding author : Woon-Hong Yeo

Abstract

The inability to objectively quantify cognitive stress in real-time with wearable devices is a crucial unsolved problem with serious negative consequences for dementia and mental disability patients and those seeking to improve their quality of life. Here, we introduce a skin-like, wireless sternal patch that captures changes in cardiac mechanics due to stress manifesting in the seismocardiogram (SCG) signals. Judicious optimization of the device's micro-structured interconnections and elastomer integration yields a device that sufficiently matches the skin's mechanics, robustly yet gently adheres to the skin without aggressive tapes, and captures planar and longitudinal SCG waves well. The device transmits SCG beats in real-time to a user's device, where derived features relate to the heartbeat's mechanical morphology. The signals are assessed by a series of features in a support vector machine regressor. Controlled studies, compared to gold standard cortisol and following the validated imaging test, show an R-squared correlation of 0.79 between the stress prediction and cortisol change, significantly improving over prior works. Likewise, the system demonstrates excellent robustness to external temperature and physical recovery status while showing excellent accuracy and wearability in full-day use.

논문정보

형식Research article
게재일2023년 12월 (BRIC 등록일 2024-01-04)
연구진국외 연구진
분야 바이오·의료융합 > 의료기기

댓글 작성 로그인

CV 열람하기

연락처 보기