Convergence analysis of dynamic fNIRS-Kalman filter for brain imaging framework

Functional near-infrared spectroscopy (fNIRS) is a non-invasive technique for detection of brain activity by measuring the absorbed quantity of near-infrared light. Absorbed light corresponds to concentration changes in oxygenated (HbO) and de-oxygenated (HbR) hemoglobins in blood during brain activity. Recently, some research is presented on real-time fNIRS-BCI applications to estimate the parameters of a GLM as states of a Kalman filter which finds applications in cortical brain imaging and brain-machine interface systems. In this study, the Lyapunov stability theory is utilized to perform the convergence analysis of dynamic fNIRS-Kalman filter for brain imaging framework. A simulation study is carried out to validate the analysis. The framework is found to be convergent in the absence of noises. The framework is robust to some unknown extent against systematic/colored and random/white noises.