Training method of two-dimensional cryoelectron microscope image denoising modeland denoising method

Abstract

The invention provides a training method of a two-dimensional cryoelectron microscope image denoising model, and the method comprises the steps: S1, obtaining an original image data set, extracting a pure noise region corresponding to each image from the original image data set to obtain pure noise samples, and forming a pure noise data set; S2, training a generative model by using the pure noise data set obtained in the step S1 to obtain a noise generation model, and generating a plurality of new pure noise samples by using the trained noise generation model to expand the pure noise data set to obtain a new pure noise data set; S3, migrating pure noise samples in the new pure noise data set to the noise-free image generated by simulation to obtain a noise-free and noise-containing data set composed of noise-free and noise-containing samples; and S4, training the convolutional neural network to convergence by using the noiseless-noisy data set obtained in the step S3.

Description

technical field [0001] The present invention relates to the field of image processing, specifically to the technical field of cryo-electron imaging, and more specifically to a training method and a denoising method for a two-dimensional cryo-electron microscope image denoising model. Background technique [0002] Cryo-electron microscopy is a widely used technique in structural biology to resolve high-resolution three-dimensional (3D) structures of proteins and polymeric complexes from a series of two-dimensional (2D) micrographs. However, since the signal-to-noise ratio (SNR) of the original cryo-EM images can only reach 0.01-0.1, which is the lowest in all imaging fields, this greatly affects the accuracy and efficiency of downstream analysis of cryo-EM images and reduces the structure determination. credibility. Therefore, image quality restoration operations are often required, followed by particle picking, structural segmentation, and other cryo-EM image analysis proce...

AI Technical Summary

Problems solved by technology

However, there are two problems with this method: (1) The framework provides a general model, which can be used directly without the need for the user to retrain the network model

However, this can limit users' use of publicly available datasets or others' datasets, as raw imaging frame sequences are often not publicly available

Images