Blood flow imaging is an essential part of biomedical research, particularly in the vascular and neurovascular physiology. It includes various imaging modalities, with laser speckle contrast imaging (LSCI) being one of the most extensively used tools for the rapid wide-field flow characterization.
For years, the capability of LSCI to become a quantitative tool has been discussed. Being based on the contrast relation to the speckle correlation time, the method requires a robust model and its correct parametrization. Main uncertainties are (i) light scattering and particle motion regimes which define the form of the field autocorrelation function g1 and (ii) static scattering and speckle averaging effects on the intensity correlation function g2. Multi-exposure laser speckle contrast imaging and proper system calibration can solve the later issue, but in order to evaluate g1 form, one has to directly measure speckle autocorrelation.
We introduce the dynamic laser speckle imaging (DLSI) as a new step in the wide-field speckle dynamics analysis. By utilizing a high-speed camera and recording backscattered light at more than 20000 frames per second we are able to measure the temporal intensity correlation (g2) in the mice cortex. We demonstrate that DLSI data can be used to estimate all parameters of the speckle autocorrelation model. By finding the best fit model for each pixel, we show that all three types of known g1 models can be found in the cortex and that the best fit model depends on the vessel size. Furthermore, we explore the commonly used model for the blood flow index and explain its deviations from the actual flow speed. We show that DLSI can be used to calibrate LSCI, thus solving contrast imaging problems and providing a lightweight quantitative tool for the blood flow imaging.