Investigation of Deep Learning Models for Predicting Diabetic Foot Ulcers in Diabetes Patients

Diabetic foot, known as Chronic diabetes complication, is widespread and can lead to a higher probability of lower limb amputation. This, in turn, results in a significant societal and economic problem. It can be prevented or it is possible to delay diabetic foot by identifying the risk factors at an early stage. By implementing the deep learning method, earlier disease examination can prevent patients from receiving more intensive preventive treatment. A systematic literature review was conducted, analyzing 15 articles that explore the use of deep learning for predicting the likelihood of developing diabetic foot ulcers. The articles were decidedly contradictory, in terms of, data usage, specificity, and accuracy. The ResNet50 v2 has gained the perfect accuracy of $\mathbf{1 0 0 \%}$ in the study. In subsequent research endeavors, it is recommended to use distinct techniques on a high number of samples. This would enable the researcher to bring a more decisive conclusion.