Neurodeblur: A novel software for fast and accurate deconvolution of large light-sheet, confocal, or bright-field microscopy stacks

Abstract

We developed NeuroDeblur, a novel software solution for deconvolving 3D microscopy data that significantly enhances the visibility of fine details—such as neurons or nerve fibers—in large image stacks captured using light-sheet, confocal, or bright-field microscopy. NeuroDeblur employs synthetic point spread functions (PSFs) tailored to each imaging modality by leveraging a sophisticated optical model of the imaging process. Additionally, experimental PSFs derived from recordings of fluorescent beads can be utilized. A newly developed module automatically extracts and averages the PSF. The software also features an advanced adaptive background removal algorithm and automatic block-wise processing of large datasets, enabling the handling of very large image stacks (>100 GB). Through extensive parallelization and GPU acceleration, NeuroDeblur achieves high-speed performance; for example, on a PC equipped with a state-of-the-art NVidia graphics board, a 3D stack comprising approximately 1 billion voxels can be deconvolved in less than 5 minutes.