Deep Clinical Imaging System for Automated Detection and Severity Classification of Diabetic Foot Ulcers

View at publisher Summarise with VIRAT
Authors: Naresh Kumar Sripada, Shikari Shabari, V. Yogiraj, G Venu Gopal, Shaik Yacoob, S Kaliappan
Cited by: 0
References: 11

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

The study proposes a high-fidel deep clinical imaging system to detect and classify Diabetic Foot Ulcers (DFUs) with high precision and is a very important solution in the context of timely clinical intervention and early diagnosis of the ulcers. Using the DenseNet121 architecture, the model proposed can have strong feature extraction and excellent gradient flow, which will make it effectively differentiate among mild, moderate, and severe stages of ulcers. The system achieves an outstanding accuracy of 99% with a standardized DFU dataset and an optimised preprocessing pipeline which beats the performance of conventional CNNs, VGG16, and models based on autoencoders. Detailed analyses with the help of confusion matrices, ROC-AUC, Precision-Recall curves, and Grad-CAM visualizations prove the reliability and interpretability of the model used in real clinical cases. The system exhibits high generalization with changing wound textures, different lighting conditions and tissue mounting, thus making it appropriate to the real-world implementation. The article brings a clinically consistent, scalable, and highly accurate deep learning solution to diabetic wound assessment and promote patient outcomes.

Fields of Study & Subjects

Diabetic Foot Ulcer Assessment and Management Artificial Intelligence in Healthcare Advanced Neural Network Applications

Publication details

Journal / source
International Conference on Image Processing and Capsule Networks (ICIPCN)
Pages
702–709
Publication date
27 January 2026
Language
EN