Fluorescence image deconvolution method and system based on depth neural network

Abstract

The invention discloses a fluorescence image deconvolution method and system based on depth neural network, The implementation of the method includes: the original image of the fluorescence microscopewith fast continuous exposure is acquired, After preprocessing, the redundancy degree of multi-frame fluorescence images on time axis is obtained. According to the relation between redundancy degreeand preset threshold, the deconvolution strategy of single-frame fluorescence image or multi-frame fluorescence image is selected to establish the energy functional of deconvolution, and the optimization of energy functional is divided into several sub-problems. Fourier transform is used to solve the de-ambiguation subproblem. For the de-noising subproblem, the variance-stabilized transform is first performed, then the de-noising is performed by the de-noising neural network, and the image is restored by the inverse transform of the variance-stabilized transform. The two subproblems are iterated alternately. By introducing the denoising neural network into the deconvolution process instead of designing the regular term manually, a clearer fluorescence image can be restored.

Description

technical field [0001] The invention belongs to the related technical fields of fluorescence microscopic imaging, digital image processing and artificial intelligence, and more specifically relates to an image processing method for fluorescence microscopic imaging observation results using deep artificial neural network, frame alignment and merging technology and variance stable transformation. Recovery method and system. Background technique [0002] With the discovery of fluorescent proteins, fluorescence microscopy imaging has become an important method for observing the life activities of living cells. The exposure process in fluorescence microscopy imaging will inevitably produce phototoxicity and photobleaching to biological samples, which will damage the activity of biological samples. In order to maintain the viability of biological samples, the exposure time and fluorescence dose need to be reduced, which leads to the low signal-to-noise ratio of the observed fluor...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the above defects or improvement needs of the prior art, the present invention provides a method and system for deconvolution of fluorescence microscopic images based on deep neural network, thereby solving the problem of blurred fluorescence image restoration effect of existing deconvolution methods. poor technical problem

Images