Significance: Diffuse correlation spectroscopy (DCS) measures cerebral blood flow non-invasively. Variations in blood flow can be used to detect neuronal activities, but its peak has a latency of a few seconds, which is slow for real-time monitoring. Neuronal cells also deform during activation, which, in principle, can be utilized to detect neuronal activity on fast timescales (within 100 ms) using DCS. Aims: We aim to characterize DCS signal variation quantified as the change of the decay time of the speckle intensity autocorrelation function during neuronal activation on both fast (within 100 ms) and slow (100 ms to seconds) timescales. Approach: We extensively modeled the variations in the DCS signal that are expected to arise from neuronal activation using Monte Carlo simulations, including the impacts of neuronal cell motion, vessel wall dilation, and blood flow changes. Results: We found that neuronal cell motion induces a DCS signal variation of . We also estimated the contrast and number of channels required to detect hemodynamic signals at different time delays. Conclusions: From this extensive analysis, we do not expect to detect neuronal cell motion using DCS in the near future based on current technology trends. However, multi-channel DCS will be able to detect hemodynamic response with sub-second latency, which is interesting for brain-computer interfaces.
Keywords: Monte Carlo; diffuse correlation spectroscopy; hemodynamics; neuronal activation; neuronal cell motion.
© 2021 The Authors.