Demirkol Denizhan, Erol Çiğdem Selçukcan, Tannier Xavier, Özcan Tuncay, Aktaş Şamil
Faculty of Engineering, Department of Computer Engineering, Aydın Adnan Menderes University, Aydın, Turkey.
Science Faculty, Department of Biology, Division of Botany & Department of Informatics, Istanbul University, İstanbul, Turkey.
Int Wound J. 2024 Jan;21(1):e14556. doi: 10.1111/iwj.14556.
Diabetic foot ulcers can have vital consequences, such as amputation for patients. The primary purpose of this study is to predict the amputation risk of diabetic foot patients using machine-learning classification algorithms. In this research, 407 patients treated with the diagnosis of diabetic foot between January 2009-September 2019 in Istanbul University Faculty of Medicine in the Department of Undersea and Hyperbaric Medicine were retrospectively evaluated. Principal Component Analysis (PCA) was used to identify the key features associated with the amputation risk in diabetic foot patients within the dataset. Thus, various prediction/classification models were created to predict the "overall" risk of diabetic foot patients. Predictive machine-learning models were created using various algorithms. Additionally to optimize the hyperparameters of the Random Forest Algorithm (RF), experimental use of Bayesian Optimization (BO) has been employed. The sub-dimension data set comprising categorical and numerical values was subjected to a feature selection procedure. Among all the algorithms tested under the defined experimental conditions, the BO-optimized "RF" based on the hybrid approach (PCA-RF-BO) and "Logistic Regression" algorithms demonstrated superior performance with 85% and 90% test accuracies, respectively. In conclusion, our findings would serve as an essential benchmark, offering valuable guidance in reducing such hazards.
糖尿病足溃疡可能会给患者带来严重后果，比如截肢。本研究的主要目的是使用机器学习分类算法预测糖尿病足患者的截肢风险。在这项研究中，我们对2009年1月至2019年9月期间在伊斯坦布尔大学医学院水下与高压医学系接受糖尿病足诊断治疗的407例患者进行了回顾性评估。主成分分析（PCA）用于识别数据集中与糖尿病足患者截肢风险相关的关键特征。因此，创建了各种预测/分类模型来预测糖尿病足患者的“总体”风险。使用各种算法创建了预测性机器学习模型。此外，为了优化随机森林算法（RF）的超参数，还采用了贝叶斯优化（BO）进行实验。对包含分类值和数值的子维度数据集进行了特征选择过程。在定义的实验条件下测试的所有算法中，基于混合方法（PCA-RF-BO）的BO优化“RF”和“逻辑回归”算法分别以85%和90%的测试准确率表现出卓越性能。总之，我们的研究结果将作为一个重要的基准，为降低此类风险提供有价值的指导。
