Retinal Artery and Vein Classification for Automatic Vessel Caliber Grading.

Kleins Lab // Publications // Nov 18 2018

PubMed ID: 30440529

Author(s): Bhuiyan A, Hussain MA, Wong TY, Klein R. Retinal Artery and Vein Classification for Automatic Vessel Caliber Grading. Conf Proc IEEE Eng Med Biol Soc. 2018 Jul;2018:870-873. doi: 10.1109/EMBC.2018.8512287.

Journal: Conference Proceedings : … Annual International Conference Of The Ieee Engineering In Medicine And Biology Society. Ieee Engineering In Medicine And Biology Society. Annual Conference, Volume 2018, Jul 2018

Automated retinal artery and vein identification is a necessity to measure their caliber automatically and to achieve high efficiency and repeatability for a large number of images. In this paper, a novel framework for retinal artery and vein classification is provided. The proposed method utilizes the vessel crossover and color intensity profile which are the most significant features for artery and vein classification. The method first extracts retinal vascular network and then identify individual blood vessels for further classification as artery or vein. We apply deep learning algorithm based segmentation method to extract the retinal vascular network. We then identify each blood vessels to measure caliber that will be used for computing the Central Retinal Artery Equivalent (CRAE) and Central Retinal Vein Equivalent (CRVE). We map the vessel network and use the individual vessel crossover information, vessel color and intensity profile to identify individual vessel segment as artery and vein. We compared automatically classified artery and vein results with a human grader which showed an accuracy of 95%. We compare our results of caliber grading against an established semi-automated caliber grading system and protocol which showed a very high correlation of 0.85 and 0.92, for CRAE and CRVE respectively.