Deep classification and segmentation model for vessel extraction in retinal images

Yicheng Wu, Yong Xia, Yanning Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

The shape of retinal blood vessels is critical in the early diagnosis of diabetes and diabetic retinopathy. Segmentation of retinal vessels, particularly the capillaries, remains a significant challenge. To address this challenge, in this paper, we adopt the “divide-and-conque” strategy, and thus propose a deep neural network-based classification and segmentation (CAS) model to extract blood vessels in color retinal images. We first use the network in network (NIN) to divide the retinal patches extracted from preprocessed fundus retinal images into wide-vessel, middle-vessel and capillary patches. Then we train three U-Nets to segment three classes of vessels, respectively. Finally, this algorithm has been evaluated on the digital retinal images for vessel extraction (DRIVE) database against seven existing algorithms and achieved the highest AUC of 97.93% and top three accuracy, sensitivity and specificity. Our comparison results indicate that the proposed algorithm is able to segment blood vessels in retinal images with better performance.

Original languageEnglish
Title of host publicationPattern Recognition and Computer Vision - First Chinese Conference, PRCV 2018, Proceedings
EditorsCheng-Lin Liu, Tieniu Tan, Jie Zhou, Jian-Huang Lai, Xilin Chen, Nanning Zheng, Hongbin Zha
PublisherSpringer Verlag
Pages250-258
Number of pages9
ISBN (Print)9783030033347
DOIs
StatePublished - 2018
Event1st Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2018 - Guangzhou, China
Duration: 23 Nov 201826 Nov 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11257 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2018
Country/TerritoryChina
CityGuangzhou
Period23/11/1826/11/18

Keywords

  • Classification and segmentation
  • Deep learning
  • Retinal vessels segmentation

Fingerprint

Dive into the research topics of 'Deep classification and segmentation model for vessel extraction in retinal images'. Together they form a unique fingerprint.

Cite this