Highly efficient eigen decomposition based statistical optical microangiography

Background: To overcome the drawbacks of the voxel-based eigen-decomposition (vED) approach to achieve the purpose of imaging blood flow in living tissue in real time. Methods: A highly efficient and practical method for contrasting in vivo blood flow by applying eigen-decomposition (ED) filter on repeated complex-valued optical coherence tomography (OCT) B-scans eigen-decomposition (bED) is proposed. We first present basic mathematics for bED. We then validate the bED through imaging cerebral blood flow in a mouse model. Results: Through evaluating signal to noise ratio, contrast and vessel connectivity, it is found that the proposed method can better contrast blood flow with drastic saving on computational power when compared with traditional ED approach where the filtering is applied on the basis of pixel by pixel or voxel by voxel. Conclusions: The bED is practically feasible to realize real time OCT angiography. In addition, the proposed ED approach is equally applicable to the operations on repeated A-scans or volumetric scans.