Identifying Metabolic Syndrome Easily and Cost Effectively Using Non-Invasive Methods with Machine Learning Models

• DOI: 10.2147/dmso.s413829
• Publication Date: 2023-07-17T05:20:05Z
• Authors: Wei Xu; Zikai Zhang; Kerong Hu; Ping Fang; Ran Li; Dehong Kong; Miao Xuan; Yang Yue; Dunmin She; Ying Xue
• Keywords: Identification

The objective of this study was to employ machine learning (ML) models utilizing non-invasive factors achieve early and low-cost identification MetS in a large physical examination population.The enrolled 9171 participants who underwent examinations at Northern Jiangsu People's Hospital 2009 2019, determine based on criteria established by the Chinese Diabetes Society. Non-invasive characteristics such as gender, age, body mass index (BMI), systolic blood pressure (SBP), diastolic (DBP) were collected used input variables train evaluate ML for identification. Several identification, including logistic regression (LR), k-nearest neighbors algorithm (k-NN), naive bayesian (NB), decision tree (DT), random forest (RF), artificial neural network (ANN), support vector (SVM).Our all showed good performance 10-fold cross-validation except SVM model. In external validation, NB model exhibited best with an AUC 0.976, accuracy 0.923, sensitivity 98.32%, specificity 91.32%.This proposed new method using models. This approach has potential serve highly sensitive, convenient, cost-effective tool large-scale screening.