Sparse decomposition deconvolution microscopy for high-speed 3-D imaging of neuronal activity

Abstract

Understanding underlying principles of neural circuits is a central goal in neuroscience. A high-speed 3-D imaging system is required to monitor the neuronal activity of whole-brain in order to study internal brain dynamics. Here we propose Sparse Decomposition Deconvolution Microscopy (SDM), a computational microscopy method for volumetric imaging of neuronal activity. The work in this thesis is composed of two parts. The first part of the thesis describes the development of SDM and the second part describes the demonstration of its capability in larval zebrafish expressing pan-neuronal GCaMP7a. SDM overcomes the major challenge of deconvolution microscopy, that 3-D deconvolution of wide-field microscopy images is an ill-posed inverse problem, by reformulating the inverse problem and applying sparse decomposition prior to deconvolution. We demonstrate the capability of SDM via in vivo imaging of neuronal activity of the whole-brain of the larval zebrafish with a volumetric field of view of 780μm×438μm×200μm at imaging rate up to 1−4.98 volumes per second. We expect the proposed microscopy to be used as an attractive tool for high-speed 3-D calcium imaging.