An experimental setup for brain activity measurement based on near infrared spectroscopy

Recently near infrared spectroscopy (NIRS) has been introduced as a recording modality in brain-computer interface applications. For a NIRS-BCI, the change of oxy-Hb and deoxy-Hb concentration related to the level of brain activity is measured and then translated to act as a control signal for an external device. In this study, a prototypical portable system for brain signal measurement with NIRS has been designed and built. The NIRS sensor consists of a pair of light emitting diodes (LED) and a pair of photodiodes (PD) attached to a flexible black fabric. The LEDs emit light at wavelengths of about 770nm and 870nm, respectively. Reflected light is recorded by photodiodes at a defined distance from the LEDs and from the prefrontal skull area of a human subject. Anoximetry circuit connected to a Beagleboard, an embedded dual-core CPU developer's board, drives the sensor. Beagleboard is utilized to generate driver signals for the oximetry circuit, the analog-to-digital conversion and the signal collection. Collected signals are sent to a PC, where the signal is processed with MATLAB based signal-processing algorithms. The signals from the subject at rest and concentration conditions are processed to observe the fluctuations and trends of the signal. Based on the results, we claim that the system has the ability to discriminate between the recorded brain signals in two conditions that could be used as a control signal for a BCI.