Changes in scalp and cortical blood flow induced by voluntary hyperventilation are investigated by near-infrared
diffusing-wave spectroscopy. The temporal intensity autocorrelation function g(<sup>2</sup>) (τ) of multiply scattered light is
recorded from the forehead of subjects during hyperventilation. Blood flow within the sampled tissue volume is
estimated by the mean decay rate of g(<sup>2</sup>) (τ) . Data measured from six subjects show that the pattern of the hemodynamic
response during 50 s hyperventilation is rather complicated: within the first 10 s, in three subjects an initial increase in
blood flow is observed; from 10 s to 20 s, the mean blood flow is smaller than its baseline value for all six subjects; for
the duration from 20 s to 30 s, the blood flow increases again. However, after 30 s the change is not consistent across
subjects. Further study on one of these subjects by using two receivers probing the blood flow in the cortex and in the
superficial layers simultaneously, reveals that during hyperventilation, the direction of change in blood flow within the
scalp is opposite to the one in the brain. This helps to understand the complicated hemodynamic response observed in
An analytical theory has been proposed to describe multiple dynamic light scattering (MDLS) in a multi-layered turbid
medium with optical and dynamical heterogeneities. To examine the validity of this theory, Monte Carlo method is used
for simulating the photon correlation diffusion in the medium. The simulation results are then compared with that of the
analytical prediction. A comprehensive investigation has been carried out, including cases of one finite layer, one
semi-infinite layer, two finite layers, and three layers with the last layer being semi-infinite. In simulations optical
parameters are varied in a large range: the absorption and the reduced scattering coefficients are taken from 0.1 to 0.7
cm<sup>-1</sup> and 2.0 to 20.0 cm<sup>-1</sup>, respectively. The Monte Carlo results, in most cases, are in an excellent agreement with that of
the analytical theory, demonstrating the effectiveness of the analytical theory for multi-layered MDLS.
Diffusion coefficients in the human sensorimotor and visual cortices were measured using diffusing-wave spectroscopy. Motor and visual activation leads to increases of the diffusion coefficients in the respective cortical areas over the values at rest.
We use near-infrared dynamic multiple scattering of light [diffusing-wave spectroscopy (DWS)] to detect the activation of the somato-motor cortex in 11 right-handed volunteers performing a finger opposition task separately with their right and left hands. Temporal autocorrelation functions g(1)(r,) of the scattered light field are measured during 100-s periods of motor task alternating with 100-s resting baseline periods. From an analysis of the experimental data with an analytical theory for g(1)(r,) from a three-layer geometry with optical and dynamical heterogeneity representing scalp, skull, and cortex, we obtain quantitative estimates of the diffusion coefficient in cortical regions. Consistent with earlier results, the measured cortical diffusion coefficient is found to be increased during the motor task, with a strong contralateral and a weaker ipsilateral increase consistent with the known brain hemispheric asymmetry for right-handed subjects. Our results support the interpretation of the increase of the cortical diffusion coefficient during finger opposition being due to the functional increase in cortical blood flow rate related to vasodilation.