Fast macula detection and application to retinal image quality assessment. (January 2020)

• Record Type: Journal Article
• Authors: Alais, Robin; Dokládal, Petr; Erginay, Ali; Figliuzzi, Bruno; Decencière, Etienne
• Is Part Of: Biomedical signal processing and control. Volume 55(2020)
• Publication Date: 2020-01
• Subjects: Image quality assessment -- Convolutional neural networks -- Retinal imaging -- Macula detection; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2019.101567
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A real-time algorithm for assessing retinal image quality is proposed. The neural network used is thousands of times smaller than state-of-the-art. The algorithm is able to accurately locate the fovea at no extra cost. Performance on a clinical database is similar to human one. The method can be used to save both patient and physician time. Abstract: In this article, we present a segmentation algorithm for assessing retinal image quality with respect to the visibility of the macular region. An image is considered of acceptable quality if the macular region is clearly visible and entirely in the field of view. Additionally, for acceptable images, the method is able to locate the fovea with a maximal error of 0.34 mm. The algorithm is based on a lightweight fully-convolutional network, several thousand times smaller than state-of-the-art networks investigated so far in preliminary studies. We obtain near-human performance for assessing macula visibility and fovea localization. The presented method can easily be embedded in tabletop or handheld retinographs, decreasing the number of ungradable images, saving both patient and physician time. It is an important step towards automatic screening of retinal pathologies, including diabetic retinopathy, which is a major global healthcare issue.