Abstract

Functional near-infrared spectroscopy (fNIRS) is a non-invasive optical technique able to measure hemodynamic response in the brain cortex. Among the different approaches the fNIRS can be based on, the time resolved one allows a straightforward relationship between the photon detection time and its path within the medium, improving the discrimination of the information content relative to the different layers the tissues are composed of. Thus absorption and scattering properties of the probed tissue can be estimated, and from them the oxy- and deoxy-hemoglobin concentration. However, an open issue in the optical imaging studies is still the accuracy in separating the superficial hemodynamic changes from those happening in deeper regions of the head and more likely involving the cerebral cortex. In fact a crucial point is the precise estimate of the time dependent pathlength spent by photons within the perturbed medium. A novel method for the calculus of the absorption properties in time domain fNIRS, based on a refined computation of photon pathlength in multilayered media, is proposed. The method takes into account the non-ideality of the measurement system (its instrument response function) and the heterogeneous structure of the head. The better accuracy in computing the optical pathlength can improve the NIRS data analysis, especially for the deeper layer. Simulations and preliminary analysis on in vivo data have been performed to validate the method and are here presented.

© 2013 SPIE