Enhancing Classification Performance of Functional Near-Infrared Spectroscopy- Brain-Computer Interface Using Adaptive Estimation of General Linear Model Coefficients.

摘要: In this paper, a novel methodology for enhanced classification of functional near-infrared spectroscopy (fNIRS) signals utilizable in two-class (motor imagery and rest; mental rotation rest) brain-computer interface (BCI) is presented. First, fNIRS corresponding to motor are acquired from the prefrontal cortex respectively afterwards, filtered remove physiological noises. Then, modelled using general linear model (GLM), coefficients which adaptively estimated least squares technique. Subsequently, multiple feature combinations were used classification. The best accuracies achieved five subjects, versus rest 79.5, 83.7, 82.6, 81.4, 84.1% whereas those 85.5, 85.2, 87.8, 84.8%, respectively, support vector machine (SVM). These results compared with obtained conventional hemodynamic response. By means proposed methodology, average accuracy was significantly higher (p < 0.05). serve demonstrate feasibility developing high-classification-performance fNIRS-BCI.