Statistical Mapping of Speckle Autocorrelation for Visualization of Hyperaemic Responses to Cortical Stimulation

Abstract

Statistically mapped speckle autocorrelation images (SAR) were used to track the hemodynamically active perfusion regions in the rat cortex during and following DC current stimulation with high transverse spatial resolution (38 um). The SAR images provided a spatio-temporal information about the net activation patterns of Cerebral Blood Flow (CBF) changes over a period of time as against those changes for each frame interval estimated using spatial contrasts derived from the first order spatial statistics. Thus the information about the relative maxima of perfusion during a Transient Hyperaemic Episode (THE) across different regions in the imaging window could be identified without the need for actually having to estimate the spatial contrast maps of the imaged region for each frame contained in the time window of observation. With the application of DC stimulation, the regions with a high correlation in the temporal fluctuations were representative of the areas that underwent least changes in activation. By varying the intensity of stimulation, THEs were observed for stimulation current densities in the range 0.1-3.8 mA/mm2 using both the derived speckle contrast maps and concurrently on a Laser Doppler Flow meter, with its probe positioned 1 mm from the site of stimulation. For current densities below the lower threshold of stimulation, the SAR images revealed an unprecedented reduction in the surge amplitude at sites distal to the region of stimulation. This was accompanied by an increase in pixel areas representing minimally active regions of perfusion ("perfusion islets") with no identifiable peak in the hemodynamic responses estimated from speckle contrast variations. The SAR images can be a useful tool for visualization of slow wave perfusion dynamics during cortical stimulation.