Extended Kalman filtering for enhancing a cyclic pattern-updating scheme of single-pixel dynamic fluorescence spatial frequency domain imaging

Abstract

A single-pixel imaging technique applied to fluorescence spatial frequency domain imaging (f-SFDI) brings many positive benefits, but its low frame rate will also lead to severe quantitative degradation when dynamically imaging a vibrant target. This work presents a novel, to the best of our knowledge, single-pixel imaging approach that combines the extended Kalman filtering (EKF) and a cyclic one-pattern updating for an enhanced dynamic f-SFDI. The cyclic one-pattern updating scheme enables the dynamic imaging at a high frame rate, and on this basis, the imaging process of an intensity temporally varying target (assuming no structure motion in the scene) is dynamically modelled, and accordingly, the surface intensities and images at each sampling time point simultaneously estimated via the EKF. Simulation and phantom validations demonstrate that the method can improve the quantitative accuracy of the results. An in vivo experiment performed on two mice for dynamic monitoring of photosensitizer doses in a photodynamic therapy further demonstrates the clinical feasibility of the proposed method.