AUTHOR=Bhatia Surbhi , Alam Shadab , Shuaib Mohammed , Hameed Alhameed Mohammed , Jeribi Fathe , Alsuwailem Razan Ibrahim 

TITLE=Retinal Vessel Extraction via Assisted Multi-Channel Feature Map and U-Net

JOURNAL=Frontiers in Public Health

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2022.858327

DOI=10.3389/fpubh.2022.858327

ISSN=2296-2565

ABSTRACT=Early detection of vessels from fundus images can effectively prevent the permanent retinal damages caused by retinopathies such as glaucoma, hyperextension, and diabetes. Concerning the red colour of both retinal vessels and background and the vessel’s morphological variations, the current vessel detection methodologies fail to segment thin vessels and discriminate them in the regions where permanent retinopathies mainly occur. This research aims to suggest a novel approach to take the benefit of both traditional template-matching methods with recent deep learning solutions. These two methods are combined in which the response of a Cauchy matched filter is used to replace the noisy red channel of the fundus images. Consequently, a U-shaped fully connected convolutional neural network (U-net) is employed to train end-to-end segmentation of pixels into vessel and background classes. To provide enough training images and speed up the training per each instance, each preprocessed image is divided into several patches. To test the proposed method,  the DRIVE public database has been analyzed,  and metrics such as Sensitivity,  Specificity,  Accuracy,  and Precision have been measured for evaluation. The evaluation indicates the average classification accuracy is up to 0.9640 on the employed dataset.