Detecção de potenciais corticais antecipatórios em sinais de eletroencefalografia (EEG) durante a condução de carros

Abstract

The recognition of the driver’s intention from electroencephalographic signals (EEG) may be useful in the development of brain computer interface (BCI) to be used in synergy with intelligent vehicles. This can be beneficial to improve the quality of interaction between the driver and the car, for example, providing a response from the smart car aligned with the intention of the driver. In this study, the anticipation is considered as the cognitive state that leads to specific actions while driving a car. Therefore, we propose to investigate the presence of anticipatory patterns in EEG signals while driving vehicles to determine two specific actions (1) left and (2) turn right, a few milliseconds before such actions take place. An experimental protocol was proposed to record EEG signals of 5 individuals as they operate a virtual reality simulator non-invasive - it was designed for this experiment - which simulates driving a virtual car. The experimental protocol is a variant of the paradigm of contingent negative variation (CNV) with Go and Nogo conditions in virtual reality training system. The results of this study indicate the presence of anticipatory patterns observed in slow cortical potentials in the time domain (medium EEG signal) and the frequency (Power Spectra and phase coherence). This opens a range of possibilities in the development of BCI systems - based on anticipatory signals - that connect the driver to the intelligent vehicle favoring a decision-making to assess the intentions of drivers may eventually prevent accidents while driving.