A Point-Spread-Function-Aware Filtered Backprojection Algorithm for Focal-Plane-Scanning Optical Projection Tomography

Traditional optical projection tomography requires the use of low numerical aperture objectives in order to achieve a large depth of field and approximate parallel projection geometry. However, low NA objectives suffer from poor resolution, resulting in blur in the reconstructed image. In this paper, we present a modified filtered backprojection method suitable for focal-plane-scanning optical projection tomography (FPSOPT), where each projection is obtained by scanning through focal planes during collection. We show that FPS-OPT has an exact inversion formula akin to a filtered backprojection, but that incorporates the system’s point-spread-function to recover a deblurred 3D volume. With simulations, we demonstrate that FPS-OPT permits the use of high numerical aperture objectives that lead to more accurate images. We further illustrate the technique on experimentally acquired data from a fluorescently-labeled zebrafish larva, which shows that our approach reduces out-of-focus blur.