DocumentCode :
3682471
Title :
A study of effect of N-back and arithmetic mental workload on performance of P300-Speller
Author :
Yuqian Chen; Yufeng Ke; Guifang Meng; Hongzhi Qi; Feng He; Peng Zhou; Dong Ming
Author_Institution :
Department of Biomedical Engineering, College of Precision Instruments and Optoelectronics Engineering, Tianjin University, China
fYear :
2015
Firstpage :
163
Lastpage :
167
Abstract :
As one of the most important brain-computer interface (BCI) systems, P300-Speller has been extensively studied and is attracting much public attention. According to user´s ability of eye movements, two kinds of P300-Speller are existed-overt attention speller and covert attention speller. P300-Speller depends on the difference between event-related potential (ERP) evoked by target stimulus and non-target stimulus to achieve character recognition and output. In practical applications of P300-Speller, users may inevitably encounter complex situations and various mental activities that have proved to affect users´ ERP. Therefore, it is suspected that mental workload will also affect the performance of P300-Speller. In the current study, the effect of mental workload on overt and covert attention P300-Speller was studied. 3-back and mental arithmetic tasks were performed based on the classical Farewell paradigm. The results under workload condition and zero-load condition were comparatively analyzed. They showed that mental workload in the forms of 3-back task and arithmetic task will lead to obviously lower recognition accuracy in both overt and covert attention pattern, attributed to the effect of mental workload on amplitude and discriminability of ERP. The impaired performance of P300-Speller caused by mental workload may be a challenge in the practical applications of P300-Speller.
Keywords :
"Accuracy","Character recognition","Support vector machines","Feature extraction","Brain-computer interfaces","Electroencephalography","Classification algorithms"
Publisher :
ieee
Conference_Titel :
Awareness Science and Technology (iCAST), 2015 IEEE 7th International Conference on
Type :
conf
DOI :
10.1109/ICAwST.2015.7314040
Filename :
7314040
Link To Document :
بازگشت