CLASSIFICATION OF FUNCTIONAL NEAR-INFRARED IMAGING BASED HEMODYNAMIC PATTERNS RECORDED AT MENTAL ARITHMET?C AND RESTING

摘要

trenFunctional near-infrared spectroscopy (fNIRS) is a non-invasive opticalimaging technique used in brain-computer interface (BCI) systems. It is used tomeasure deoxyhemoglobin and oxyhemoglobin proportions that occur during aspecific activity in the brain region (motor and visual activity, auditorystimulus, etc.). In this study, hemodynamic patterns were recorded from 8participants during mental arithmetic and rest activities. Features have beenextracted for this by using detrended fluctuation analysis, entropy and Hjorthparameters methods. The distinctive feature vectors obtained after the featureselection process have been applied to support vector machines (SVM),multilayer artificial neural networks (MLANN) and k-nearest neighbors (k-NN)classifiers. As a result, the best classification accuracy was 97.17% when SVMclassifier was used.