Depth-varying deconvolution of light-sheet microscope images

The light-sheet microscope is a powerful instrument for 3D imaging of biological samples. These samples are typically cleared using chemical methods. However, even cleared samples can still produce image blur, which increases with depth. We propose a computational method to correct this depth-dependent blur. First, we calculate the uniform point spread function (PSF) of the microscope. Next, we adjust its shape along the sample depth to reconstruct a set of depth-varying PSFs. Finally, we use these PSFs for nonuniform deconvolution of the 3D image. The method was successfully tested on both simulated and real data.