Detection of macular diseases in optical coherence tomography image. Issue 3 (3rd May 2020)

• Record Type: Journal Article
• Authors: Liu, Xiaoming; Yang, Zhou; Hu, Wei; Liu, Jun; Zhang, Kai
• Is Part Of: International journal of parallel, emergent and distributed systems. Volume 35:Issue 3(2020)
• Page Start: 260
• Page End: 272
• Publication Date: 2020-05-03
• Subjects: Optical coherence tomography -- local binary pattern -- histogram of oriented gradient -- random forest; Parallel computers -- Periodicals; Electronic data processing -- Distributed processing -- Periodicals; Computer algorithms -- Periodicals; 004.35
• DOI: 10.1080/17445760.2018.1472261
• Languages: English
• ISSNs: 1744-5760

Abstract

Abstract : Vision decreasing and blindness are great threats to health, which can be caused by various kinds of macular diseases. Optical coherence tomography (OCT) is widely used in the diagnosis of macular disease. Ophthalmologists often need to check and analyse lots of OCT images and then give eye disease report for each image, which is time-consuming and inefficient. Thus, the automated classification method is necessary. In this paper, we proposed an approach to determine the status (normal or abnormal) of retina by OCT images. The proposed method can differentiate several kinds of macular diseases from normal image, including macular edema, macular hole and age-related macular degeneration. Besides the geometry features, our method takes advantage of two texture feature (local binary pattern histograms and histogram of oriented gradient) to improve the accuracy rate. The experimental results on a public available dataset demonstrate that our approach is effective and achieve higher accuracy (all accuracy >0.98, AUC can reach 1). Flowchart of the proposed detection method. GRAPHICAL ABSTRACT: