Separation of the hemodynamic responses in a simultaneous audio-stimuli using ICA: An fNIRS study

The `cocktail party effect´ has the ability to focus on one audio stimulus while interfered from other stimuli or noise. It will be interesting to extract multi-response in human auditory cortex for multi audio-stimuli. Functional near-infrared spectroscopy (fNIRS) technique is used to measure hemoglobin changes in the auditory cortex, when a subject listens to the multiple audio-stimuli. As a silent imaging technique, the fNIRS is most suitable tool for monitoring the response from auditory cortices in human. The independent component analysis (ICA) is one of the most exciting algorithms for signal processing and it is used to separate the different stimuli. Each step of the pre-processing, statistical analysis, and evaluate the results are carefully analyzed. Here we present the results that separate the hemodynamic response to two different audio-stimuli obtained by ICA. Also, we evaluate the result by investigating the t-value and correlation with the desired and original signals, respectively.