Authors:
D. Sumathi, Kuruva Sree Sowmya, Dhivya Kalaiselvan, R. Kesavamoorthy, Madhan Raj Gopi Akila, Jaime Alfonso Flores Navas
Addresses:
Department of Computer Science and Engineering, CMR Institute of Technology, Bengaluru, Karnataka, India. Department of Computer Science, University of Stuttgart, Stuttgart, Baden-Württemberg, Germany. Faculty of Sciences, National Autonomous University of Mexico, Coyoacán, Mexico.
Abstract:
To develop a robust and effective design, the proposed research uses a model that analyses blood vessels in retinal images by leveraging Deep Learning and Image Processing techniques. Deep learning, one of the AI techniques, has experienced rapid growth and is widely used in the medical field to analyse medical images, driving advancements in technology and medicine. The U-Net semantic segmentation neural network is the primary deep learning model used in our study. Training data consists of a set of retinal images, and the model’s performance is assessed on the training set. Once testing is complete, the results help evaluate the accuracy and effectiveness of the model and the techniques utilised in this research. This study indicates that deep learning segmentation techniques can offer valuable insights beyond standard pixel-based detection techniques. The proposed method gives a better understanding of retinal blood vessel analysis. Researchers used the DRIVE dataset for this research. It has 20,512x512 training and testing images for the model’s training and testing. The neural network's blood vessel shape recognition helps diagnose eye diseases. Diseases include hypertension, diabetic retinopathy, etc. The neural network's effectiveness makes the system suited for real-time blood vessel detection. This paper accurately mapped major and minor blood arteries. Thus, 96.53% of this research succeeded. The mean intersection over union is 8184. This exceeds the previous E-NET model of 7744.
Keywords: Blood Vessel Segmentation; Deep Learning (DL); Diagnosis of Retinal Diseases; Digital Image Recognition; Retinal Vasculature Analysis; Retinal Vessel Segmentation.
Received on: 19/01/2025, Revised on: 11/05/2025, Accepted on: 20/08/2025, Published on: 05/01/2026
DOI: 10.64091/ATIHL.2026.000262
AVE Trends in Intelligent Health Letters, 2026 Vol. 3 No. 1 , Pages: 17-26