Optical doppler coherence tomography algorithms: Quantitative analysis

Abstract

We present, to our knowledge, the first quantitative comparisons of five flow estimation algorithms for Optical Doppler Coherent Tomography. We have used simulated ID, 2D, and in vivo blood flow data to investigate the estimation accuracy and sensitivity of these algorithms. We introduce, among these five methods, 3 new algorithms: weighted centroid, sliding-window filtering, and autocorrelation. The weighted centroid, sliding-window filtering and the existing adaptive centroid algorithms have considerably higher sensitivity and accuracy over the autocorrelation and existing cross-correlation methods.