Enhancement of Medical Imaging Technique for Diabetic Retinopathy: Realistic Synthetic Image Generation Using GenAI

Damodharan Palaniappan¹, Tan Kuan Tak², K. Vijayan³, Balajee Maram⁴, Pravin R Kshirsagar⁵, Naim Ahmad^6,*
1 Department of Information Technology, Marwadi University, Rajkot, 360003, India
2 Engineering Cluster, Singapore Institute of Technology, Singapore, 828608, Singapore
3 Electronics and Communication Engineering Department, Sapthagiri NPS University, Bangalore, 560057, India
4 School of Computer Science and Artificial Intelligence, SR University, Warangal, 506371, India
5 Electronics Telecommunication Engineering, J D College of Engineering Management, Nagpur, 441501, India
6 College of Computer Science, King Khalid University, Abha, 61421, Saudi Arabia
* Corresponding Author: Naim Ahmad. Email: email

Computer Modeling in Engineering & Sciences https://doi.org/10.32604/cmes.2025.073387

Received 17 September 2025; Accepted 11 November 2025; Published online 04 December 2025

Download PDF

Abstract

A phase-aware cross-modal framework is presented that synthesizes UWF_FA from non-invasive UWF_RI for diabetic retinopathy (DR) stratification. A curated cohort of 1198 patients (2915 UWF_RI and 17,854 UWF_FA images) with strict registration quality supports training across three angiographic phases (initial, mid, final). The generator is based on a modified pix2pixHD with an added Gradient Variance Loss to better preserve microvasculature, and is evaluated using MAE, PSNR, SSIM, and MS-SSIM on held-out pairs. Quantitatively, the mid phase achieves the lowest MAE (98.76 ± 42.67), while SSIM remains high across phases. Expert review shows substantial agreement (Cohen’s κ = 0.78–0.82) and Turing-style misclassification of 50%–70% of synthetic images as real, indicating strong perceptual realism. For downstream DR stratification, fusing multi-phase synthetic UWF_FA with UWF_RI in a Swin Transformer classifier yields significant gains over a UWF_RI-only baseline, with the full-phase setting (Set D) reaching AUC = 0.910 and accuracy = 0.829. These results support synthetic UWF_FA as a scalable, non-invasive complement to dye-based angiography that enhances screening accuracy while avoiding injection-related risks.

Keywords

Diabetic retinopathy; synthetic image generation; GenAI; medical imaging; ultra-widefield retinal imaging; enhanced medical imaging datasets; multi-scale structural similarity

Downloads
- Full-Text PDF
Citation Tools
- BibTex
- EndNote
- RIS

67

View
13

Download
1

Like

Mu-Net: Multi-Path Upsampling Convolution Network for Medical Image Segmentation
Jia Chen, Zhiqiang He, Dayong...
An Effective Machine-Learning Based Feature Extraction/Recognition Model for Fetal Heart Defect Detection from 2D Ultrasonic Imageries
Bingzheng Wu, Peizhong Liu, Huiling...
An Interpretable CNN for the Segmentation of the Left Ventricle in Cardiac MRI by Real-Time Visualization
Jun Liu, Geng Yuan, Changdi Yang,...
SepFE: Separable Fusion Enhanced Network for Retinal Vessel Segmentation
Yun Wu, Ge Jiao, Jiahao Liu
An Improved Lung Cancer Segmentation Based on Nature-Inspired Optimization Approaches
Shazia Shamas, Surya Narayan Panda,...

Enhancement of Medical Imaging Technique for Diabetic Retinopathy: Realistic Synthetic Image Generation Using GenAI

Abstract

Keywords

67

13

1

Further Information

Guidelines

Follow Us

Join Us

Contact Us

WhatsApp:

Share Link