{Hw Vol. 9, No. 1, pp. 27-33 (2007)



Autocorrelation for Verification of Task Dependence and Cross Correlation for Separation of Heart Pulse Wave in NIRS Signal using Optical Topography to BCI


I. Kurimoto, K. Murakami, S. Ooeda, M. Nakagawa and S. Tachi


Computer interface is useful for nerve intractable disease patient. It is the most important that development of brain-computer inferface(BCI ) which is direct communication of brain information and external world. Near-Infrared Spectoroscopy(NIRS) for cerebral function analysis was developed in recent years. The measuring method and the analysis procedure are not yet established to BCI. In this paper, we will propose autocorrelation analysis method for task and cerebral activation, and cross correlation analysis method for separation of heart pulse wave using the optical topography. As a result of autocorrelation to the NIRS signal, it found out that it is effective in separation of task component, and noise reduction. We next experimented in auditory selective attention, and analyzed the autocorrelation function dependence of task. Comparing the selective attention with none selective attention, it found out that increasing the autocorrelation signal at the auditory se-lective attention. Moreover, using the proposed cross correlation method, when performing the cross correlation of a unit impulse(time periods for 0.8sec) and NIRS signal (sampling 25msec), it succeeded in separation of heart pulse wave.


Key Words : Autocorrelation, Cross correlation, NIRS, Optical topography, Task




{Hw : The Japan Society for Welfare Engineering