Preliminary results on P300 detection using machine learning when modulating task reaction time
Su-Kyoung Kim, Elsa Andrea Kirchner
In Proceedings of the 18th Annual Meeting of the Organization for Human Brain Mapping, (OHBM-2012), 10.6.-14.6.2012, Bejing, o.A., Jun/2012.
Zusammenfassung (Abstract)
:
In recent years, machine learning (ML) method has been applied to detect cognitive states, e.g. the event-related potential (ERP) P300 has been often used for BCI applications. The P300 correlates to cognitive processing elicited by successful recognition of important messages. Unlike most BCIs (e.g. P300 speller), the application in the robotic field requires to detect more naturally evoked ERP activity during multiple tasking. We developed a scenario requiring a main task (e.g. manipulating) and a secondary task (recognition of important warnings).
Previous results proved a successful P300 detection in such complex scenarios. However, in a real scenario, it is not always possible to react to important information immediately. The goal of the study was to investigate the performance of P300 detection during delayed reaction to warnings.
Files:
120412_Preliminary_results_on_P300_detection_using_machine_learning_when_modulating_task_reaction_time_OHBM_Kim.pdf
