Classifying the Levels of Fear by Means of Machine Learning Techniques and VR in a Holonic-Based System for Treating Phobias. Experiments and Results

Abstract

This paper presents the conceptual design, implementation and evaluation of a VR based system for treating phobias that simulates stress-provoking real-world situations, accompanied by physiological signals monitoring. The element of novelty is the holonic architecture we propose for the real-time adaptation of the virtual environment in response to biophysical data (heart rate (HR), electrodermal activity (EDA) and electroencephalogram (EEG)) recorded from the patients. In order to enhance the impact of the therapy, we propose the use of gamified scenarios. 4 acrophobic patients have been gradually exposed to anxiety generating scenarios (on the ground and at the first, 4th and 6th floors of a building, at different distances from the railing), where EEG, EDA and HR have been recorded. The patients also reported their level of fear on a scale from 0 to 10. The treatment procedure consisted in a VR-based game where the subjects were exposed to the same heights. They had to perform some small quests at various distances from the railing and report the in-game stress level, while biophysical data was recorded. The real-life scenarios have been repeated, with the purpose of assessing the efficiency of the VR treatment plan.