A rib extraction method and device based on deep learning

Abstract

The embodiment of the present invention provides a method for extracting ribs based on deep learning, comprising the following steps: training sample preprocessing; establishing a first deep convolutional neural network for rib extraction, and performing training on the first deep convolutional neural network through the training samples training; establishing a second deep convolutional neural network for discriminating and optimizing the first deep convolutional neural network, optimizing the first deep convolutional neural network; extracting medical The rib image in the image. The first embodiment of the present invention adopts the method of sample augmentation to improve the training effect of the model; the second is to use a deep convolutional neural network to achieve a higher recognition rate; the third is to consider the boundary when performing feature fusion. Loss and area loss are more conducive to the convergence of the network; fourth, in the training part of the network, the model's attention to difficult samples is increased, and the network can converge better.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, and in particular relates to a rib extraction method and device based on deep learning. Background technique [0002] Medical image analysis is a common method used by doctors when diagnosing lung diseases. In actual work, on the one hand, ribs often block part of the lung tissue, which affects the doctor's accurate diagnosis of the disease. In order to overcome this problem, the ribs in the medical image can be separated in advance, so that the ribs can be removed from the medical image, so as to form a medical image that only shows the lung tissue structure, and reduce the adverse effect of ribs on disease diagnosis. . On the other hand, the extracted rib image can provide a basis for analyzing rib anatomy and diagnosing various related diseases. However, the existing rib extraction work usually adopts the method of manual marking. First, this method mainly takes up a lot of...

AI Technical Summary

Problems solved by technology

However, the existing rib extraction work usually adopts the method of manual marking. First, this method mainly takes up a lot of time for doctors. Second, the accuracy of the marked rib area is low. Once the marking is wrong, it is difficult to find and correct it in time.

Images