“…This is because the characteristics of the visual stimuli provide the framework to interact with the system, to select targets, and, consequently, to evoke P300 responses. Indeed, parameters, such as shape, size, color, and type of flash, may enhance or diminish the difference between target and non-target responses and, thus, may influence the performance of the machine-learning model [ 20 , 21 , 22 ]. Therefore, an alternative to improve the accuracy and reliability of P300-BCIs is to employ stimuli properties that evoke stronger P300 responses [ 23 ] and that might also elicit other ERP components that occur before or after the P300 [ 24 ].…”