Biomarkers of Anxiety and Depression – A Novel Method of Tracking the Autonomic Nervous System with Machine Learning

This study presents a novel method to understand biomarkers of anxiety and depression by using a concept called Sympathetic Transition Points (STP), which is indicative of Autonomic Nervous System (ANS) dynamics. Wearables-based Electrodermal Activity (EDA) and Blood Volume Pulse (BVP) data were collected from 61 controls, 60 individuals with depressive symptoms, and 110 individuals with anxiety. By monitoring ANS activity, patterns related to anxiety and mood states and their transitions in real-time were identified, using machine learning. Analysis revealed clear distinctions between groups and enabled tracking of mental state changes. A score of .99 F1, with an ROC of 1 was achieved in automatically classifying anxiety, depression, and neutral states with this method. The method lays the groundwork for automated mental health assessments in real-world settings, introducing an efficient and objective screening protocol. By utilizing wearable technology, machine learning, and ANS monitoring, this work advocates for improved early detection and intervention strategies in mental healthcare.