A Meaningful Machine Learning Model for Predicting Amputation Rate of Patients with Diabetic Foot Ulcer

Abstract

Background. Diabetic foot (DF) disease, which includes ulcers, infections and gangrene of the feet, is one of the leading causes of disability worldwide. Due to the high disability rate and expensive treatment cost of diabetic foot, doctors and patients all hope to forecast the prognosis in time and give early intervention. With the development of artificial intelligence technology, more and more methods are used in the diagnosis and prognosis prediction of chronic diseases. Machine learning, a type of artificial intelligence, has excellent predictive effects with a certain accuracy.¹ The results of diabetic foot are affected by many factors, so it is necessary for the machine learning to reasonably predict the relationship between input variables and output variables, and to correct and tolerate faults.² Objective. To develop an accurate and applicable predictive model for diabetic foot amputation and use it to guide clinical diagnosis and treatment, indicating the direction for the prevention of diabetic foot amputation. Methods and Materials. This retrospective study collected the basic data of 150 patients with DFU who met the study criteria in Beijing Shijitan Hospital from January 2019 to December 2022. Above all, We divided them into amputation group and non-amputation group based on prognostic outcome. Then we used Lasso algorithm to screen relevant risk factors, and predictive models were built with support vector mechanism(SVM) to input risk factors and predict amputation. Besides, we divided the test set and training set by 5-fold cross-validation. The area under the receiver operating characteristic (ROC) curves of the model were 0.89. This model’s calibration capability was 19.614 through Hosmer-Lemeshow test (p=0.012). Conclusion. In summary, our survey data suggested that C-reactive protein (CRP) in the infection index and the Wagner scale of the affected foot might play a vital role in predicting diabetic foot amputation. The predictive model we constructed can accurately estimate the rate of amputation during hospitalization in DFU patients. In addition, the model allows for personalized analysis of patients' risk factors.