Machine learning models for predicting metabolic dysfunction-associated steatotic liver disease prevalence using basic demographic and clinical characteristics

Journal of Translational Medicine, Journal Year: 2025, Volume and Issue: 23(1)

Published: March 28, 2025

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a global health concern that necessitates early screening and timely intervention to improve prognosis. The current diagnostic protocols for MASLD involve complex procedures in specialised medical centres. This study aimed explore the feasibility of utilising machine learning models accurately screen large populations based on combination essential demographic clinical characteristics. A total 10,007 outpatients who underwent transient elastography at First Affiliated Hospital Gannan Medical University were enrolled form derivation cohort. Using eight characteristics (age, educational level, height, weight, waist hip circumference, history hypertension diabetes), we built predictive (classified as none or mild: controlled attenuation parameter (CAP) ≤ 269 dB/m; moderate: 269-296 severe: CAP > 296 dB/m) employing 10 algorithms: logistic regression (LR), multilayer perceptron (MLP), extreme gradient boosting (XGBoost), bootstrap aggregating, decision tree, K-nearest neighbours, light machine, naive Bayes, random forest, support vector machine. These externally validated using National Health Nutrition Examination Survey (NHANES) 2017-2023 datasets. In hospital outpatient cohort, algorithms demonstrated robust capabilities. Notably, LR achieved highest accuracy (ACC) 0.711 test cohort 0.728 validation coupled with areas under receiver operating characteristic curve (AUC) values 0.798 0.806, respectively. Similarly, MLP XGBoost showed promising results, achieving an ACC 0.735 registering AUC 0.798. External NHANES datasets yielded consistent (0.831), (0.823), (0.784) performing robustly. constructed can general population. approach significantly enhances feasibility, accessibility, compliance provides effective tool large-scale assessments strategies.

Language: Английский

The role of Triglyceride Glucose-Waist Circumference (TyG_WC) in predicting metabolic dysfunction-associated steatotic liver disease among individuals with hyperuricemia

BMC Gastroenterology, Journal Year: 2025, Volume and Issue: 25(1)

Published: April 4, 2025

The incidence of metabolic dysfunction-associated steatotic liver disease (MASLD) among individuals with hyperuricemia is significantly high. aim this study was to identify effective biomarkers for the detection MASLD patients hyperuricemia. We conducted an analysis involving 3424 participants from National Health and Nutrition Examination Survey (1999-2020). To potential significant variables, we employed Boruta's algorithm, SHapley Additive exPlanations (SHAP) random forests. Multivariable logistic regression models were utilized assess odds ratio (OR) developing MASLD. evaluate accuracy clinical value our prediction model, receiver operating characteristic (ROC) curves decision curve (DCA) curves. Among population (mean [SD] age, 54 [20] years, 1788 [52.22%] males) hyperuricemia, 1670 had Using SHAP forests, suggested that Triglyceride Glucose-Waist Circumference (TyG_WC) one most variables in predicting risk, area under (AUROC) 0.865. restricted spline (RCS) revealed a positive association between TyG_WC MASLD, when compared lowest quantile TyG_WC, risk highest 137.96 times higher. predictive strategy incorporating notably enhanced threshold probabilities spanning approximately 0% 100%, resulting improvement net benefit. Our found

Language: Английский