Quantitative measurement of blood flow speed using probabilistic analysis of OCT speckle

We propose a novel approach to use OCT speckle analysis to quantify absolute flow speed in skin blood vessels. Using a bespoke scan acquisition pattern, we empirically compute a probability density function for the L1-norm decorrelation measure as a function of displacement. During clinical scanning, we calculate the decorrelation value within a small window at each point. By inverting the probability function, we are able to calculate a probability distribution over the possible values of blood displacement, and hence flow speed. We demonstrate the algorithm on 10 human patients who have suffered heart failure and 10 age-matched healthy control subjects.