Year: 2023
Type: dissertation
Abstract: This research focused on the development and evaluation of methods to study stress-related physiology using wearable sensors in two populations: individuals with major depressive disorder (MDD) and depressive symptoms, and adolescents with autism spectrum disorder (ASD). The research employed electrodermal activity (EDA), skin temperature (ST), and heart rate (HR) measured through electrocardiography (ECG) and photoplethysmography (PPG). Two types of time frames were considered: momentary and day-based. In the momentary approach, short segments of physiological data were analyzed in relation to labeled outcomes. In the day-based approach, prolonged trends in physiological data were examined. The research involved evaluating state-of-the-art EDA feature extraction methods, quality indication tools, and analyzing baseline levels, events, and circadian rhythms of EDA, HR, and ST. Existing datasets were used for stress and depression analysis, while a new dataset was collected for studying stress in autistic adolescents. The main findings include the development of a novel flow for EDA data processing and the successful use of wearables to identify unique stress-related physiological patterns in subclinical populations.
Source: Lirias
Author: Lutin, Erika
Language: English
Topic: Emotion and Mood Recognition
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6999872263