Lipids & metabolic liver disease
Machine Learning–Based Biomarker Identification for Early Diagnosis of Metabolic Dysfunction–Associated Steatotic Liver Disease
Authors' conclusion
These findings indicate that assessing a variety of biomarkers, across demographic, metabolic, lipid, and standard biochemistry categories, may provide valuable initial insights for diagnosing patients for MASLD and hepatic fibrosis.
Study context
- Population
- Participants of the National Health and Nutritional Examination Survey (2017–2020) with FibroScan® hepatic steatosis and stiffness measurements
- Participants
- 15,560 participants
- Exposure
- Steatosis model vs Stiffness model
- Outcome
- Prediction of MASLD and hepatic fibrosis by Random Forest models (the most accurate)
Methodological detailsmachine learning (Random Forest), propensity score matching 1:1 for age and sex, logistic regression, sensitivity analysis using XGBoost and Recursive Feature Elimination
- Machine learning (Random Forest)
- Propensity score matching 1:1 for age and sex
- Logistic regression
- Sensitivity analysis using XGBoost and Recursive Feature Elimination
Main finding
As reported in the abstract
| Group | Accuracy | P | Specificity | Sensitivity |
|---|---|---|---|---|
| Steatosis model | 79.59% | P < .001 | 84.65% | |
| Stiffness model | 86.07% | P < .001 | 98.01% |
Accuracy
Steatosis model79.59%
Stiffness model86.07%
P
Steatosis modelP < .001
Stiffness modelP < .001
Specificity
Steatosis model84.65%
Stiffness model— not stated in the abstract
Sensitivity
Steatosis model— not stated in the abstract
Stiffness model98.01%
Observational data: associations, not effects.
Original publication
Machine Learning–Based Biomarker Identification for Early Diagnosis of Metabolic Dysfunction–Associated Steatotic Liver Disease. J Clin Endocrinol Metab. Published 21 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →