Objective detection of Steady State Visual Evoked Responses for different lighting conditions

Abstract

This work aims at evaluating the performance of two Objective Response Detection (ORD) techniques: the Magnitude-Squared Coherence (MSC) and the Multiple Coherence (MC) for detecting Steady State Visual Evoked Response (SSVEP) for two different lighting conditions (scotopic and photopic). EEG signals were collected (using a wireless headset) from ten volunteers without history of neurological diseases during stimulation with white LED flickering at frequencies of 6, 7, 8, 9, 10 and 11 Hz. The experimental protocol was composed by 13 sessions in a dark room and 13 in a bright one. The frequency order was randomized. The MSC was applied to derivations [P7], [P8], [O1] and [O2], and MC to the combinations of leads [O1-O2], [O1-O2-P7], [O1-O2-P8] and [O1-O2-P7-P8]. Both estimates were calculated using M = 11 epochs. The detection rates for MSC presented an important decrease at photopic condition compared to the scotopic one for all derivations and all frequencies. The employment of MC to [O1-O2-P7-P8] allowed overcoming this limitation, leading to detection percentages of at least 80% for all frequencies (except 9 Hz). These results show that MC is a promising technique for building SSVEP Bran Computer Interfaces (BCI).