時系列解析によるNIRSデータの信号処理

抄録

Recently, Near-infrared spectroscopy (NIBS) is employed in various research fields as the method of evaluating brain function. However, an effective methodology specific to those researches has not yet been established. For instance, methods such as comparison of scatter chart of acquired data, determination of degree of change in hemoglobin levels, and judgment of the topography image, are performed. These solving methods pose problems influencing the difference in the measurement, the difference in the hemoglobin movement each the humans, and the propriety of the experimental content for the research. Therefore, authors focus on the time series data obtained from the NIBS and attempt to employ trend analysis, which is one of the effective methods for this solving, and apply the frequency analysis via the autoregressive e(AR)model and the moving average method for smoothing data, aiming to eliminate heartbeat signal with noise data and the treatment afterwards. Consequently, the proposal of a highly effective method was made by attempting an analysis method that had not yet been established. Furthermore, the programming for the NIIZS analysis software is scheduled to be performed after resolution of the various problems associated with it.