EDA-based Data Stream Pattern Analysis and Peak Detection Algorithm for Substance Users

Wireless communication is rapidly expanding to biomedical and health areas due to 5G networks and technology improvements. This paper continues our previous studies in real time substance use detection to enable future near-real-time behavioral intervention in natural environments. Specifically, this project explores the unclarified but critical electrodermal activity (EDA) patterns from streaming data collected using wireless biosensors. Our unique peak detection method tailored to cocaine users focuses on EDA data streams and correlated bio-physiological signals collected from skin temperature, locomotion, and related effects visible in data streams.