Screening Diabetic Foot Ulcer using Artificial Intelligence Modelling based on Digital Image Analysis: A Systematic Review

Abstract

Aims: This study conducted a systematic review with the aim of identifying the predictive models used in the development of AI-based digital image analysis in Diabetic Foot Ulcer cases and determining the features and segmentation used in the construction of Diabetic Foot Ulcer screening algorithm models. Methods: A systematic review was conducted by searching articles from ScienceDirect, PubMed, ProQuest, and CINAHL databases using a combination of relevant keywords. The selection process followed the PRISMA guidelines and article quality was assessed using the Mixed Methods Appraisal Tool (MMAT). Results: A search of the electronic databases produced 374 research articles within the time range of 2019–2024, with an average article quality of 93% (strong). The results of this systematic review show that out of the eight articles, all were involved in developing an AI model, with seven journals developing convolutional neural network models and one journal developing an artificial neural network model. Digital image analysis involved colour segmentation of tissues and areas of Diabetic Foot Ulcer, which can be used for screening. Conclusion: The convolutional neural network AI model was used in two-dimensional digital imaging modalities for patients with diabetic foot ulcers. The development of an accurate prediction model can provide an automated system for assessing and monitoring Diabetic Foot Ulcer.