Particle swarm optimization-based convolutional neural network identification method for pulmonary nodules

Abstract

The invention belongs to the field of combining artificial intelligence and medical image analysis, and relates to a method for identifying pulmonary nodules based on a convolutional neural network optimized by a particle swarm algorithm, comprising the following steps: obtaining a CT slice image of a patient's lung; Carry out preprocessing, label the obtained CT slice images to form a data set; construct a convolutional neural network; use particle swarm optimization to optimize the hyperparameters of the convolutional neural network; use the marked data set to optimize the convolutional neural network Perform training; use the trained convolutional neural network to extract pulmonary nodule features. The invention optimizes the convolutional neural network through the particle swarm algorithm, solves the problem of difficult manual selection of the optimal hyperparameters of the convolutional neural network, and has a good recognition effect on pulmonary nodules.

Description

technical field [0001] The invention belongs to the field of combining artificial intelligence and medical image analysis, and relates to a method for identifying pulmonary nodules based on a convolutional neural network optimized by a particle swarm algorithm. Background technique [0002] At present, computed tomography (CT) is a commonly used technology for effectively screening early lung cancer. The development of CT technology has significantly improved the detection rate of early lung cancer. Compared with conventional X-ray photography, we can see higher-resolution images of lung anatomy under CT; but for radiologists, observing and interpreting these images is a complex and tedious task. This is because in complex CT images, pulmonary nodules appear similar to some lung structures, are low-density and small in size, and may be located very close to or closely related to blood vessels or lung boundaries. The borders are connected. Second, another factor is the larg...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the deficiencies of the prior art, the present invention proposes a method for identifying pulmonary nodules based on the convolutional neural network optimized by the particle swarm optimization algorithm. It is difficult to manually select the problem, and it has a good recognition effect on pulmonary nodules.

Images