A New Machine Learning-Based Complementary Approach for Screening of NAFLD (Hepatic Steatosis)

Area of Science:

• Hepatology
• Medical Informatics
• Machine Learning

Background:

• Non-Alcoholic Fatty Liver Disease (NAFLD) is a leading cause of chronic liver disease worldwide.
• Early detection of NAFLD is crucial for mitigating disease progression, reducing mortality, and enhancing patient longevity.
• Current clinical approaches often prioritize disease management over early screening and detection.

Purpose of the Study:

• To develop and evaluate machine learning-based intelligent models for the early screening of Hepatic Steatosis (Non-alcoholic Fatty Liver).
• To utilize liver functionality and physiological parameters for improved NAFLD detection.
• To create gender-specific models to enhance screening accuracy.

Main Methods:

• Development of gender-specific machine learning models for Non-Alcoholic Fatty Liver Disease screening.
• Application of customized data processing techniques.
• Utilizing publicly available population data from the National Health and Nutrition Examination Survey III (NHANES-III).

Main Results:

• The developed models achieved maximum sensitivities of approximately 72% for males and 71% for females.
• Maximum specificities were recorded at 74% for males and 75% for females.
• A comparative analysis of different model performances was conducted.

Conclusions:

• Machine learning models show promise for the early screening of Non-Alcoholic Fatty Liver Disease.
• Gender-specific models may offer improved accuracy in detecting hepatic steatosis.
• Further research can refine these models for enhanced clinical utility in NAFLD detection.