376 0

Mental stress analysis based on ultra-short-term HRV with time-frequency method

Title
Mental stress analysis based on ultra-short-term HRV with time-frequency method
Author
이승재
Advisor(s)
김인영
Issue Date
2022. 8
Publisher
한양대학교
Degree
Doctor
Abstract
In modern society, mental stress is the main cause of many diseases. According to the World Health Organization (WHO), the social effects associated with stress are increasing rapidly and have been affecting the lives of men and women of all ages. There are two types of stress: acute stress and chronic stress. Acute stress influences the sympathetic nervous system response belonging to the activity of the autonomic nervous system, which is also called the fight or flight response. If anyone continues to face acute stressful situations for a long period of time, chronic stress that causes mental and physical problems occurs due to imbalance of the autonomic nervous system. Mental stress can be quantitatively analyzed using heart rate variability (HRV), which can analyze the autonomic nervous system response, including the sympathetic and parasympathetic nervous system. Several studies have reported that assess acute mental stress based on HRV. Most of the stress assessment studies conducted a short-term HRV analysis using 5-minute data. However, as the capability of wearable devices increases, it is necessary to develop an intelligent system that can measure and evaluate acute stress in less than 5 minutes. In response to necessity, there are several studies that conducted ultra-short-term HRV analysis using an HRV signal of less than 5 minutes. Previous studies compared the results of ultra-short-term HRV analysis with those of short-term HRV analysis, but the results of each study were different. In addition, they proceeded without considering changes in stress response over time in acute stress situations. In this study, we propose new HRV parameters using the Empirical mode decomposition (EMD), a time-frequency method and evaluated feasibility through mental stress analysis. The study was conducted in three steps as follows. First, we present IMF-energy parameters by decomposing Heart rate variability (HRV) through the Empirical mode decomposition (EMD) method suitable for ultra-short-term HRV analysis. Second, the results of ultra-short-term HRV analysis of time lengths (1-min, 2-min, 3-min, 4-min) were compared with the short-term one considering the stress response change in the stress situation and the weight value of the proposed EMD-Energy parameters in the classification model were compared. Finally, the Perceived Stress Scale (PSS) questionnaire, which can measure perceived chronic stress, was divided into a group with a high–PSS score and low-PSS score, and HRV parameters in the resting state and stress state were compared using statistical analysis. As a result, we confirmed the appropriateness of the proposed IMF-Energy parameter because there was no significant difference between classification accuracy and frequency domain parameters in short-term HRV analysis. Ultra-short-term HRV analysis confirmed the change of stress response in a stress situation. The accuracy of each time length was compared based on the result that the stress response decreased over time in a stress situation. The results showed that the highest classification accuracy was obtained with 88.8% of the 3-min segment, and the first 2-min segment showed that it was higher than the result of the short-term HRV analysis. Based on the results, the optimal time length for stress assessment using ultra-short-term HRV data is 3 minutes, but a minimum length of 2 minutes is suggested. In addition, we confirmed the feasibility of EMD-Energy parameters in ultra-short-term HRV analysis by comparing the weight values for each HRV parameter. Finally, as a result of the PSS questionnaire, the HRV parameters of parasympathetic tone showed lower values in the group with high PSS than low PSS in the resting state. Also, through ultra-short-term HRV analysis, it was shown that there was a difference between groups in all states.
URI
http://hanyang.dcollection.net/common/orgView/200000627907https://repository.hanyang.ac.kr/handle/20.500.11754/174344
Appears in Collections:
GRADUATE SCHOOL OF BIOMEDICAL SCIENCE AND ENGINEERING[S](의생명공학전문대학원) > BIOMEDICAL ENGINEERING(생체의공학과) > Theses (Ph.D.)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE