Abstract:
This research presents the development of a cost effective functional near infrared spectroscopy (fNIRS) equipment for monitoring human brain activity in the pre frontal cortex region. Combination of three dual wavelength near infrared LED sources and eight photodiodes form a 12 channel brain activity acquisition module. The two wavelengths 760nm and 850nm are emitted one by one after regular intervals and the acquired data is continuously logged into an excel file. In this design intensities falling onto the photodiodes are fed into the Modified Beer Lambert Law (MBLL). Using this law we observe the behaviour of oxygenated hemoglobin (HbO) and deoxygenated hemoglobin (HbR) at various instants. Change in concentration of HbO and HbR according to human brain activity is recorded. Feature extraction is performed using heuristic methods and Linear Discriminant Analysis (LDA) has been used to verify, one the appropriateness of selected feature and second the validation of data acquired through the developed fNIRS equipment based on minimum effective criteria of classification accuracy for distinguishing two or more tasks.
