Netiwit Kaongoen1, Sungho Jo2. 1. School of Computing, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea. Electronic address: netiwit.k@cs.kaist.ac.kr. 2. School of Computing, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea. Electronic address: shjo@kaist.ac.kr.
Abstract
BACKGROUND: Brain-computer interface (BCI) is a technology that provides an alternative way of communication by translating brain activities into digital commands. Due to the incapability of using the vision-dependent BCI for patients who have visual impairment, auditory stimuli have been used to substitute the conventional visual stimuli. NEW METHOD: This paper introduces a hybrid auditory BCI that utilizes and combines auditory steady state response (ASSR) and spatial-auditory P300 BCI to improve the performance for the auditory BCI system. The system works by simultaneously presenting auditory stimuli with different pitches and amplitude modulation (AM) frequencies to the user with beep sounds occurring randomly between all sound sources. Attention to different auditory stimuli yields different ASSR and beep sounds trigger the P300 response when they occur in the target channel, thus the system can utilize both features for classification. RESULTS: The proposed ASSR/P300-hybrid auditory BCI system achieves 85.33% accuracy with 9.11 bits/min information transfer rate (ITR) in binary classification problem. COMPARISON WITH EXISTING METHODS: The proposed system outperformed the P300 BCI system (74.58% accuracy with 4.18 bits/min ITR) and the ASSR BCI system (66.68% accuracy with 2.01 bits/min ITR) in binary-class problem. The system is completely vision-independent. CONCLUSIONS: This work demonstrates that combining ASSR and P300 BCI into a hybrid system could result in a better performance and could help in the development of the future auditory BCI.
BACKGROUND: Brain-computer interface (BCI) is a technology that provides an alternative way of communication by translating brain activities into digital commands. Due to the incapability of using the vision-dependent BCI for patients who have visual impairment, auditory stimuli have been used to substitute the conventional visual stimuli. NEW METHOD: This paper introduces a hybrid auditory BCI that utilizes and combines auditory steady state response (ASSR) and spatial-auditory P300 BCI to improve the performance for the auditory BCI system. The system works by simultaneously presenting auditory stimuli with different pitches and amplitude modulation (AM) frequencies to the user with beep sounds occurring randomly between all sound sources. Attention to different auditory stimuli yields different ASSR and beep sounds trigger the P300 response when they occur in the target channel, thus the system can utilize both features for classification. RESULTS: The proposed ASSR/P300-hybrid auditory BCI system achieves 85.33% accuracy with 9.11 bits/min information transfer rate (ITR) in binary classification problem. COMPARISON WITH EXISTING METHODS: The proposed system outperformed the P300 BCI system (74.58% accuracy with 4.18 bits/min ITR) and the ASSR BCI system (66.68% accuracy with 2.01 bits/min ITR) in binary-class problem. The system is completely vision-independent. CONCLUSIONS: This work demonstrates that combining ASSR and P300 BCI into a hybrid system could result in a better performance and could help in the development of the future auditory BCI.
Authors: Tiago Zanotelli; Antonio Mauricio Ferreira Leite Miranda de Sá; Eduardo Mazoni Andrade Marçal Mendes; Leonardo Bonato Felix Journal: Med Biol Eng Comput Date: 2019-08-09 Impact factor: 2.602