Neural Reactivity to Haptics: Virtual Tasks Versus Physical Tasks

This study investigates the effectiveness of haptic feedback in hand rehabilitation exercises, within both virtual reality (VR) and real-world settings, to enhance upper limb functionality in post-stroke recovery. We developed a VR system that incorporates haptic feedback into pinch simulations and compared its efficacy with similar real-world pinch training exercises, primarily focusing on the rehabilitation of the paretic hand. Our main objective was to ascertain the similarity of the simulated haptic feedback in the VR environment to actual tactile feedback in the real world. This was achieved by analyzing electroencephalogram (EEG) features and psychometric data collected from 35 healthy subjects. A key aspect of our approach was the use of machine learning (ML) techniques to classify the EEG signal features as originating from VR or Real tasks. This classification was based on data from the top-n electrode channels, enabling us to assess the degree of similarity in brain responses to tactile stimulation in both environments. The study's results revealed that the brain activity patterns elicited by haptic feedback were comparable in both the VR and real-world setups. Additionally, the VR feedback system was positively received by users, scoring high on experience and satisfaction metrics. This research provides a replicable methodology for assessing tactile feedback in existing rehabilitation systems and lays the groundwork for the development of novel therapeutic applications. Our findings are crucial for designing future rehabilitation interventions that effectively utilize haptic feedback, potentially leading to improved treatment outcomes for stroke patients.