Scholarlink

Background <p>Diabetic foot ulcers (DFUs) are a severe complication affecting diabetic patients, potentially leading to lower limb amputations if not detected and treated promptly.</p> Objective <p>This review explores advanced techniques, particularly deep learning models, for&#xa0;non-invasive DFU detection and monitoring.</p> Methods <p>The review examines various aspects, including plantar thermography, artificial intelligence (AI), and deep learning algorithms such as convolutional neural networks (CNNs). It discusses innovative solutions like the Intelligent Diabetic Assistant (IDA), Dia-Shoe, and Smart Diabetic Foot Ulcer Scoring System (ScoreDFUNet). The QUADAS-2 assessment is used to evaluate the reliability of included studies across four domains: patient selection, index test, reference standard, and flow/timing.</p> Results <p>The review identifies key research gaps, including the need for diverse datasets, real-time implementation, multi-modal data integration, clinical validation, and improved interpretability. It highlights the potential of advanced techniques in identifying ulcers and grading their severity.</p> Conclusion <p>Addressing identified challenges and leveraging AI and deep learning can improve early detection, treatment, and management of DFUs. Future directions include developing user-friendly technologies, creating comprehensive databases, and exploring deep learning models for clinical application, ultimately reducing complications and improving patient outcomes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A comprehensive review of advanced techniques that aid in development of non-invasive diagnostic tool for diabetic foot ulcer detection and monitoring using deep learning models

  • 期刊论文

  • 发表日期：

    2025年11月13日

查看全文
  • M. Kayalvizhi,
  • D. Maheswari

摘要

Background

Diabetic foot ulcers (DFUs) are a severe complication affecting diabetic patients, potentially leading to lower limb amputations if not detected and treated promptly.

Objective

This review explores advanced techniques, particularly deep learning models, for non-invasive DFU detection and monitoring.

Methods

The review examines various aspects, including plantar thermography, artificial intelligence (AI), and deep learning algorithms such as convolutional neural networks (CNNs). It discusses innovative solutions like the Intelligent Diabetic Assistant (IDA), Dia-Shoe, and Smart Diabetic Foot Ulcer Scoring System (ScoreDFUNet). The QUADAS-2 assessment is used to evaluate the reliability of included studies across four domains: patient selection, index test, reference standard, and flow/timing.

Results

The review identifies key research gaps, including the need for diverse datasets, real-time implementation, multi-modal data integration, clinical validation, and improved interpretability. It highlights the potential of advanced techniques in identifying ulcers and grading their severity.

Conclusion

Addressing identified challenges and leveraging AI and deep learning can improve early detection, treatment, and management of DFUs. Future directions include developing user-friendly technologies, creating comprehensive databases, and exploring deep learning models for clinical application, ultimately reducing complications and improving patient outcomes.