Kenji Kato1, Kensho Takahashi2, Nobuaki Mizuguchi3, Junichi Ushiba4. 1. Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan; Department of Rehabilitation Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku, Tokyo, 160-8582, Japan. 2. Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan. 3. Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan; The Japan Society for the Promotion of Science, 5-3-1 Kojimachi, Chiyoda-ku, Tokyo, 102-0083, Japan. 4. Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan; Keio Institute of Pure and Applied Sciences (KiPAS), Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa, 223-8522, Japan. Electronic address: ushiba@brain.bio.keio.ac.jp.
Abstract
BACKGROUND: Neurofeedback of event-related desynchronization (ERD) in electroencephalograms (EEG) of the sensorimotor cortex (SM1) using a brain-computer interface (BCI) paradigm is a powerful tool to promote motor recovery from post-stroke hemiplegia. However, the feedback delay attenuates the degree of motor learning and neural plasticity. NEW METHOD: The present study aimed to shorten the delay time to estimate amplitude modulation of the motor-imagery-related alpha and beta SM1-ERD using a lock-in amplifier (LIA) algorithm. The delay time was evaluated by calculating the value of the maximal correlation coefficient (MCC) between the time-series trace of ERDs extracted by the online LIA algorithm and those identified by an offline algorithm with the Hilbert transform (HT). RESULTS: The MCC and delay values used to estimate the ERDs calculated by the LIA were 0.89±0.032 and 200±9.49ms, respectively. COMPARISON WITH EXISTING METHOD(S): The delay time and MCC values were significantly improved compared with those calculated by the conventional fast Fourier transformation (FFT), continuous Wavelet transformation (CWT), and autoregressive (AR) algorithms. Moreover, the coefficients of variance of the delay time and MCC values across trials were significantly lower in the LIA compared with the FFT, CWT, and AR algorithms. CONCLUSIONS: These results indicate that the LIA improved the detection delay, accuracy, and stability for estimating amplitude modulation of motor-related SM1-ERD. This would be beneficial for BCI paradigms to facilitate neurorehabilitation in patients with motor deficits.
BACKGROUND: Neurofeedback of event-related desynchronization (ERD) in electroencephalograms (EEG) of the sensorimotor cortex (SM1) using a brain-computer interface (BCI) paradigm is a powerful tool to promote motor recovery from post-stroke hemiplegia. However, the feedback delay attenuates the degree of motor learning and neural plasticity. NEW METHOD: The present study aimed to shorten the delay time to estimate amplitude modulation of the motor-imagery-related alpha and beta SM1-ERD using a lock-in amplifier (LIA) algorithm. The delay time was evaluated by calculating the value of the maximal correlation coefficient (MCC) between the time-series trace of ERDs extracted by the online LIA algorithm and those identified by an offline algorithm with the Hilbert transform (HT). RESULTS: The MCC and delay values used to estimate the ERDs calculated by the LIA were 0.89±0.032 and 200±9.49ms, respectively. COMPARISON WITH EXISTING METHOD(S): The delay time and MCC values were significantly improved compared with those calculated by the conventional fast Fourier transformation (FFT), continuous Wavelet transformation (CWT), and autoregressive (AR) algorithms. Moreover, the coefficients of variance of the delay time and MCC values across trials were significantly lower in the LIA compared with the FFT, CWT, and AR algorithms. CONCLUSIONS: These results indicate that the LIA improved the detection delay, accuracy, and stability for estimating amplitude modulation of motor-related SM1-ERD. This would be beneficial for BCI paradigms to facilitate neurorehabilitation in patients with motor deficits.