New speckle analysis method for optical coherence tomography signal based on autocorrelation

Abstract

Optical Coherence Tomography (OCT) is a noninvasive imaging technique with high resolution widely used for in vivo applications. Nonetheless, OCT is prone to speckle, a granular noise that degrades the OCT signal. Speckle statistics may, nevertheless, reveal information regarding the scatterers from which it originates. This fact is exploited by techniques such as Speckle Variance-OCT (SVOCT). SVOCT, however, doesn’t provide quantitative information, which is a major drawback for the use of speckle based techniques on OCT. In the present work we attack this problem, proposing a new method for analysis of speckle in OCT signal, based on autocorrelation. We associate the changes in decorrelation time of the signal with the changes in flow velocity. It is expected that greater velocities result in lower decorrelation times. To verify that, milk was pumped through a microchannel at different velocities, and the decorrelation time was computed for a single point in the center of the microchannel, sampled at 8 kHz rate. Our results suggest that for flows rates greater than 1 μl/min it is possible to associate decorrelation time with flow velocity, while velocities below that value are not distinguishable, supposedly due to the Brownian motion. For flow rates above 50 μl/min our acquisition rate doesn’t get enough sampling information, as the decorrelation time gets too low. These results indicate that Speckle based techniques may be used to get quantitative information of flow in OCT samples, which can be used to assist in many diagnostics modalities, as well as map such flow regions.