Deep Learning Architectures in Medical Imaging - Architectures for the Early Detection of Diabetic Complications

Read full text Summarise with VIRAT
Authors: Adan Khan
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This presentation, delivered at the International Webinar Series on Healthcare Technologies and Medical Sciences (IWS-HTMS), addresses the critical "Workflow Gap" in current diabetic screening. Current clinical practices often rely on detecting visually evident macroscopic lesions, which can fundamentally delay necessary therapeutic interventions. With a projected diabetic population of 600 million, manual ophthalmoscopy and wound assessment are becoming increasingly unscalable due to high inter-grader variability and extreme time constraints. The research explores how Deep Learning Architectures, specifically the shift from local filters (CNNs) to global self-attention mechanisms (Vision Transformers), can exceed human visual perception. By analysing multi-dimensional pixel interactions, these models can mathematically correlate latent feature spaces with sub-clinical microvascular degradation.

Fields of Study & Subjects

Retinal Imaging and Analysis Advanced Neural Network Applications AI in cancer detection

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
14 March 2026
Language
EN