Cervical cell classification method based on space, channel and scale attention fusion learning

Abstract

The invention relates to a cervical cell classification method based on space, channel and scale attention fusion learning. The method comprises the following steps: preparing a training sample; constructing a channel attention module; constructing a space attention module; constructing a scale attention module; constructing a deep network based on space, channel and scale attention fusion learning; constructing a cervical cell image classifier; and predicting the category of the image: loading a network structure and weight parameters of the deep network based on space, channel and scale attention fusion learning, and inputting the cervical cell image into the deep network based on space, channel and scale attention fusion learning to obtain a classification result. According to the invention, a classification model capable of classifying five types of cervical cell images is constructed, the cervical cell images are classified through the method, doctors can be assisted in analysis, and the burden of pathologists can be relieved; and medical resource contradictions can be solved, small hospitals such as grassroots and villages can be covered, and the national overall screening level can be improved.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a cervical cell classification method based on spatial, channel and scale attention fusion learning. Background technique [0002] Cytological examination is the most commonly used method for screening early cervical cancer. A cervical smear usually contains tens of thousands of cervical cells. The screening process brings a great burden to pathologists, and fatigue occurs from time to time. . Computer-aided analysis technology establishes a pattern recognition model based on the characteristics of tumor cells to automatically analyze cell smears, and uses objective evaluation criteria to improve screening efficiency, reduce false negative rates, and reduce the burden of film reading for pathologists. [0003] The attention mechanism extracts the weight distribution from the features, and then applies the weight distribution to the original features to change th...

AI Technical Summary

Problems solved by technology

For the problem of cervical cell classification, there are few studies on the combination of attention mechanism and convolutional neural network. In particular, the deep learning method using channel, space and scale information fusion to solve the problem of cervical cell classification has not been reported yet.

Images