Three-dimensional particle category detection method and system based on convolutional neural network

Abstract

The invention provides a three-dimensional particle category detection method and system based on a convolutional neural network. The method comprises the following steps: constructing a three-dimensional mixed-scale dense convolutional neural network comprising a mixed-scale three-dimensional extended convolutional layer, dense connection and a loss function, training the convolutional neural network by using a three-dimensional frozen electron tomography image marked with the particle coordinates to obtain a particle selection model, and training the convolutional neural network by using thethree-dimensional frozen electron tomography image marked with the particle category to obtain a particle classification model; acquiring the three-dimensional frozen electron tomography image through a sliding window to obtain to-be-detected three-dimensional reconstructed subareas, predicting each subarea through the particle selection model, and combining prediction results of the subareas toobtain coordinates of each particle in the three-dimensional frozen electron tomography image; and extracting a three-dimensional image of each particle according to the coordinate of each particle, and inputting the three-dimensional image of each particle into the particle classification model to obtain the category of each particle.

Description

technical field [0001] The invention belongs to the technical field of structural biology cryo-electron tomography, and in particular relates to a three-dimensional particle type detection method and system based on a convolutional neural network. Background technique [0002] Cryo-electron tomography combined with sub-region averaging technology can obtain higher resolution in situ structures of biological macromolecules. One of the key steps is to select a large number of three-dimensional particles from electron tomography reconstruction. The current methods for selecting three-dimensional particles are divided into There are two types of manual selection and automatic selection. [0003] Some software packages provide hand-picked views from electron tomographic 3D reconstructions. The usual practice is to select the center point of the particle on the projection plane perpendicular to the z axis, and further mark the three-dimensional coordinates of the particle in the ...

AI Technical Summary

Problems solved by technology

[0012] The purpose of the present invention is to solve the problem of lack of automation in picking and classifying three-dimensional particles in cryo-electron tomography reconstruction, and the problem of too many model parameters in existing deep learning-based methods

Images