Scattering reduced imaging of spheroids via refractive index estimation and shift-variant deconvolution.

Abstract

Confocal laser scanning microscopy (CLSM) is widely used in biological research, but imaging the deep regions of three-dimensional samples like spheroids is challenging due to scattering. We propose a computational method that estimates the refractive index distribution from CLSM images, calculates position-dependent point-spread functions (PSFs) using a multi-diffraction propagation model for both excitation and emission light, and applies shift-variant deconvolution. This approach enables the resolution of deep spheroid structures that could not be resolved in conventional CLSM images. It requires no hardware modifications to conventional CLSM systems, enabling high-quality three-dimensional imaging of scattering samples using conventional equipment.