Complex decorrelation-based OCT angiography (OCTA) has the potential for quantitively monitoring hemodynamic activities. To improve the dynamic range and uncertainty for quantification, an adaptive spatial-temporal (ST) kernel was proposed. The ensemble size in decorrelation computation was enlarged by collecting samples in the spatial/ temporal dimensions. The spatial sub-kernel size was adaptively changed to suppress the bulk motion influence by solving a maximum entropy model. The improvement of dynamic range and uncertainty were validated by theoretical analyzation, numerical simulation, and in vitro/ in vivo experiments. Furthermore, proved by the in vivo experiments, the adaptive ST-kernel can also improve the separability between different stimuli and allow a reliable temporal analysis of the hemodynamic response.
|