Park, Sangin Jihyeon Ha Kim, Laehyun 2024-01-12T02:33:16Z 2024-01-12T02:33:16Z 2023-03-14 2023-01 1534-4320 https://pubs.kist.re.kr/handle/201004/75847 The phenomena of brain-computer interface-inefficiency in transfer rates and reliability can hinder development and use of brain-computer interface technology. This study aimed to enhance the classification performance of motor imagery-based brain-computer interface (three-class: left hand, right hand, and right foot) of poor performers using a hybrid-imagery approach that combined motor and somatosensory activity. Twenty healthy subjects participated in these experiments involving the following three paradigms: (1) Control-condition: motor imagery only, (2) Hybrid-condition I: combined motor and somatosensory stimuli (same stimulus: rough ball), and (3) Hybrid-condition II: combined motor and somatosensory stimuli (different stimulus: hard and rough, soft and smooth, and hard and rough ball). The three paradigms for all participants, achieved an average accuracy of 63.60± 21.62%, 71.25± 19.53%, and 84.09± 12.79% using the filter bank common spatial pattern algorithm (5-fold cross-validation), respectively. In the poor performance group, the Hybrid-condition II paradigm achieved an accuracy of 81.82%, showing a significant increase of 38.86% and 21.04% in accuracy compared to the control-condition (42.96%) and Hybrid-condition I (60.78%), respectively. Conversely, the good performance group showed a pattern of increasing accuracy, with no significant difference between the three paradigms. The Hybrid-condition II paradigm provided high concentration and discrimination to poor performers in the motor imagery-based brain-computer interface and generated the enhanced event-related desynchronization pattern in three modalities corresponding to different types of somatosensory stimuli in motor and somatosensory regions compared to the Control-condition and Hybrid-condition I. The hybrid-imagery approach can help improve motor imagery-based brain-computer interface performance, especially for poorly performing users, thus contributing to the practical use and uptake of brain-computer interface. English Institute of Electrical and Electronics Engineers Improving Performance of Motor Imagery-Based Brain-Computer Interface in Poorly Performing Subjects Using a Hybrid-Imagery Method Utilizing Combined Motor and Somatosensory Activity Article 10.1109/tnsre.2023.3237583 1 IEEE Transactions on Neural Systems and Rehabilitation Engineering, v.31, pp.1064 - 1074 IEEE Transactions on Neural Systems and Rehabilitation Engineering 31 1064 1074 Y scie scopus 000934975600005 Engineering, Biomedical Rehabilitation Engineering Rehabilitation Article ERROR POTENTIAL DETECTION CLASSIFICATION CORTEX COMMUNICATION MOVEMENTS FEATURES SINGLE HAND Somatosensory Training Performance evaluation Task analysis Protocols Electroencephalography Brain-computer interfaces BCI inefficient brain-computer interface motor imagery motor imagery training somatosensory stimuli