MSGANet-RAV: A multiscale guided attention network for artery-vein segmentation and classification from optic disc and retinal images

A Z M Ehtesham Chowdhury; Graham Mann; William Huxley Morgan; Aleksandar Vukmirovic; Andrew Mehnert; Ferdous Sohel

doi:10.1016/j.optom.2022.11.001

MSGANet-RAV: A multiscale guided attention network for artery-vein segmentation and classification from optic disc and retinal images

J Optom. 2022;15 Suppl 1(Suppl 1):S58-S69. doi: 10.1016/j.optom.2022.11.001. Epub 2022 Nov 14.

Authors

A Z M Ehtesham Chowdhury¹, Graham Mann¹, William Huxley Morgan², Aleksandar Vukmirovic², Andrew Mehnert², Ferdous Sohel³

Affiliations

¹ School of Information Technology, Murdoch University, 90 South Street, Murdoch, WA 6150, Australia.
² Lions Eye Institute, 2 Verdun Street, Nedlands, WA 6009, Australia; Centre for Ophthalmology and Visual Science, The University of Western Australia, 35 Stirling Highway, Perth, WA 6009, Australia.
³ School of Information Technology, Murdoch University, 90 South Street, Murdoch, WA 6150, Australia. Electronic address: f.sohel@murdoch.edu.au.

Abstract

Background: Retinal and optic disc images are used to assess changes in the retinal vasculature. These can be changes associated with diseases such as diabetic retinopathy and glaucoma or induced using ophthalmodynamometry to measure arterial and venous pressure. Key steps toward automating the assessment of these changes are the segmentation and classification of the veins and arteries. However, such segmentation and classification are still required to be manually labelled by experts. Such automated labelling is challenging because of the complex morphology, anatomical variations, alterations due to disease and scarcity of labelled data for algorithm development. We present a deep machine learning solution called the multiscale guided attention network for retinal artery and vein segmentation and classification (MSGANet-RAV).

Methods: MSGANet-RAV was developed and tested on 383 colour clinical optic disc images from LEI-CENTRAL, constructed in-house and 40 colour fundus images from the AV-DRIVE public dataset. The datasets have a mean optic disc occupancy per image of 60.6% and 2.18%, respectively. MSGANet-RAV is a U-shaped encoder-decoder network, where the encoder extracts multiscale features, and the decoder includes a sequence of self-attention modules. The self-attention modules explore, guide and incorporate vessel-specific structural and contextual feature information to segment and classify central optic disc and retinal vessel pixels.

Results: MSGANet-RAV achieved a pixel classification accuracy of 93.15%, sensitivity of 92.19%, and specificity of 94.13% on LEI-CENTRAL, outperforming several reference models. It similarly performed highly on AV-DRIVE with an accuracy, sensitivity and specificity of 95.48%, 93.59% and 97.27%, respectively.

Conclusion: The results show the efficacy of MSGANet-RAV for identifying central optic disc and retinal arteries and veins. The method can be used in automated systems designed to assess vascular changes in retinal and optic disc images quantitatively.

Keywords: Artery-vein classification; Deep learning; Fundus image; Guided attention; Optic disc image.

MeSH terms

Arteries
Glaucoma* / diagnosis
Humans
Optic Disk* / diagnostic imaging
Retina
Retinal Vessels / diagnostic imaging