Fundus image blood vessel segmentation method based on full convolutional neural network multi-scale features

Abstract

The invention discloses a fundus image blood vessel segmentation method based on full convolutional neural network multi-scale features. The fundus image blood vessel segmentation method comprises thefollowing steps of 1) preprocessing a fundus retina image; 2) segmenting the preprocessed image into image blocks for data expansion; 3) constructing a convolutional neural network model, and performing network training by using the expanded data; and 4) testing the trained model to obtain a segmentation result. According to the invention, one encoding structure and two different decoding structures are connected; the method provided by the invention can overcome the defects of low segmentation precision and the like caused by a small number of blood vessel image data sets and low image quality by adopting multiple skipping connections, more fully fuses the characteristics of different depths, effectively alleviates the problem of gradient disappearance caused by network depth increase, and has higher accuracy and higher robustness.

Description

Technical field [0001] The invention relates to the fields of medical image processing and computer vision, in particular to a method for segmenting blood vessels in fundus images based on convolutional neural networks. Background technique [0002] The fundus retinal capillary network is the same as human features such as faces, fingerprints, and palm prints. It also has individual uniqueness. The distribution, orientation, thickness, and curvature of the retinal blood vessels of each person are different, so the fundus information can be used for biometric identification. In addition, among all the deeper microvascular circulation networks in the human body, the only one that can be directly observed non-invasively is the retinal microvascular network. Normally, the vascular network remains unchanged for a long time, but cardio-cerebrovascular diseases such as diabetes and hypertension can cause changes in the structure and shape of the retinal microvascular network. However, ...

AI Technical Summary

Problems solved by technology

Unsupervised segmentation methods do not require prior label information, but have high requirements for image quality, usually require more additional conditions, and the segmentation effect is slightly worse

The supervised method is mainly based on the extracted feature classifier to achieve the purpose of identifying blood vessels and background. The neural network model needs to extract image features layer by layer, and it is easy to lose a lot of useful information, resulting in the neural network not being able to segment blood vessel images well.

Images