Detection of Static, Dynamic, and No Tactile Friction Based on Non-linear dynamics of EEG Signals: A Preliminary Study

Abstract

Touching an object leads to a frictional interaction between the skin and the object. There are two kinds of friction: the first contact that leads to static friction and the dragging phase that leads to dynamic friction. No study has been performed to show the effect of friction type on EEG signals. The main goal of the current study is to investigate the effect of tactile friction on non-linear features of EEG signals. Participants performed a tactile task that each of its trials had three states: the sensation of 1) static friction, 2) dynamic friction, and 3) no friction. During the experiment, EEG signals were recorded, and different linear and non-linear EEG indices were extracted and analyzed to find the effect of the tactile friction on EEG signals. Linear features such as spectral features were not a good choice to distinguish between the states. However, non-linear features such as Lyapunov exponent, Higuchis dimension, and Hurst exponent had the potential to separate the mentioned states. Results also showed signs of predictability (negative Lyapunov exponent) in the signals recorded during dynamic friction and the existence of long-range dependency (memory) in EEG signals recorded during all states. The complexity of the tactile system in Theta band was also higher than the Delta band. The results of this research not only increase our knowledge about brain non-linear dynamics in response to tactile friction but also lead to a design of a preliminary system that can automatically detect friction between the skin and surfaces.