Fluorescence to color feature-based image registration for medical augmented reality

In recent years, augmented reality with fluorescence imaging has attracted significant attention for computer aided surgeries. Overlaying functional information acquired from fluorescence images on the anatomical information of color images offers intuitive surgical guidance. In this study, we investigate the feasibility of SIFT, SURF, BRISK, and ORB algorithms in registration of fluorescence and color reflectance images for intraoperative augmented reality and surgical guidance. We have further demonstrated the potential of a new registration method combining SURF, BRISK and ORB which showed improved registration accuracy. Several sets of experiments were performed on phantoms, biological tissues with fluorescently labeled targets to evaluate the performance of various algorithms. We have demonstrated the feasibility of applying feature-based registration methods in fluorescence-to-color image registration for intraoperative augmented reality. We have also demonstrated the potential of a new featurebased image registration method combining BRISK, SURF, and ORB.