High-precision retinal blood vessel segmentation based on a multi-stage and dual-channel deep learning network.

Researchers

Journal

Modalities

Models

Abstract

The high-precision segmentation of retinal vessels in fundus images is important for the early diagnosis of ophthalmic diseases. However, the extraction for microvessels is challenging due to their characteristics of low contrast and high structure complexity. Although some works have been developed to improve the segmentation ability in thin vessels, they are just successful in recognizing the small vessels with relative high contrast.Therefore, we develop a deep learning (DL) framework with multi-stage and dual-channel network model (MSDC_NET) to further improve the thin-vessel segmentation with low contrast. Specifically, an adaptive image enhancement strategy combining multiple pre-processing and DL method is firstly proposed to elevate the contrast of thin vessels; then, a two-channel model with multi-scale perception is developed to implement the whole- and thin-vessel segmentation; and finally, a series of post-processing operations are designed to extract more small vessels in the predicted maps from thin-vessel channel.Experiments on DRIVE, STARE and CHASE_DB1 demonstrate the superiorities of the proposed MSDC_NET on extracting more thin vessels in fundus images, and the quantitative evaluations on several parameters based on the advanced ground truth further verify the advantages of our proposed DL model. Compared with the previous multi-branch method, the specificity and F1 score are improved about 2.18%, 0.68%, 1.73% and 2.91%, 0.24%, 8.38% on the three datasets, respectively.This work may provide richer information to ophthalmologists for diagnosis and treatment of vascular-related ophthalmic diseases.© 2024 Institute of Physics and Engineering in Medicine.