Rapid estimations of intensity standard deviations for optical coherence tomography angiography

Optical coherence tomography angiography (OCTA) can map microvascular networks and quantify blood flow velocities with high resolution by calculating intensity standard deviations of time-series signals. However, statistical calculations of the standard deviations need much processing time and reduce the analysis efficiency. In this study, we proposed three optimized OCTA algorithms incorporating rapid estimations of the intensity standard deviations, including the range algorithm, the mean absolute error algorithm and the maximum absolute error algorithm. The abilities of the optimized algorithms to quantify the flow velocities were validated by a flow phantom. After a rat cerebral cortex was imaged, the optimized OCTA algorithms were compared with the conventional relative standard deviation algorithm in the metrics of imaging quality and processing time. The results show that the optimized algorithms incorporating rapid estimations of the intensity standard deviations have faster processing speeds with equivalent image quality.