Retinal blood vessel segmentation using a deep learning method based on modified U-NET model

Sanjeewani,Arun Kumar Yadav,Mohd Akbar,Mohit Kumar,Divakar Yadav
DOI: https://doi.org/10.1007/s11042-024-18696-w
IF: 2.577
2024-03-16
Multimedia Tools and Applications
Abstract:Retinal blood vessel segmentation is important for detection of several highly prevalent, vision-threatening diseases such as diabetic retinopathy. Automatic retinal blood vessel segmentation is crucial to overcome the limitations posed by diagnoses by doctors. In recent times, deep learning-based methods have achieved great success in automatically segmenting retinal blood vessels from images. In this paper, a U-Net-based architecture is proposed to segment the retinal blood vessels from fundus images of the eye. Three pre-processing algorithms are proposed to enhance the performance of the proposed method further. Based on experimental evaluation of the publicly available DRIVE dataset, the proposed method achieves 0.9577 average accuracies (Acc), 0.7436 sensitivity (Se), 0.9838 specificities (Sp) and 0.7931 F1-score. The proposed method outperforms the recent state-of-art approaches in the literature.
computer science, information systems, theory & methods,engineering, electrical & electronic, software engineering
What problem does this paper attempt to address?