The Potential Value and Impact of Diagnostic Biomarkers for MAFLD Using Machine Learning Methods

The Potential Value and Impact of Diagnostic Biomarkers for MAFLD Using Machine Learning Methods

This is a case-control study that aims to build a predictive model for MAFLD based on machine learning.

Metabolic dysfunction-associated fatty liver disease (MAFLD) also known as non-alcoholic fatty liver disease (NAFLD), is one of the most prevalent liver diseases worldwide with high prevalence and economic burden, which affects 25% of global adult population. Despite extensive research on understanding the inner pathophysiology of MAFLD, it still keep growing with no approval therapy. Therefore, preventive measures are particularly important in diagnosing MAFLD. So far the liver biopsy is still the gold standard for diagnosis of MAFLD, however considering the invasive process and potential risks, it still has low acceptance for asymptomatic patients, thus non-invasive methods are necessary for this reason.

The purpose of this study is to establish a prediction model to identify MAFLD patients, which can accurately predict whether the participants have MAFLD according to the relevant metabolic indicators of the participants, without the need for invasive examinations such as tissue biopsy.

Inclusion Criteria:

1. aged 18 to 75 years;
2. meeting the diagnostic criteria of MAFLD;
3. no other organic lesions were found in imaging examination;
4. willing and able to sign informed consent.

Exclusion Criteria:

1. significant drinking history (weekly alcohol consumption ≥ 140g for male, or weekly alcohol consumption ≥ 70g for female);
2. presence of evidence for having hepatic steatosis, viral hepatitis, history of hepatic cancer, drug-induced liver injury, liver cirrhosis and other liver and biliary tract diseases;
3. major organ malfunction, severe systemic illnesses, mental health issues, or inability to complete examination;
4. pregnant or pregnancy planning female;
5. missing of important clinical data.