Hybrid intelligent systems for liver disease prediction: a demographic-aware machine learning framework

Area of Science:

• Medical Informatics
• Machine Learning in Healthcare
• Computational Biology

Background:

• Liver disease is a significant global health issue, often diagnosed late due to limitations in traditional methods.
• Current diagnostic tools for liver disease can be invasive, expensive, and not widely accessible.

Purpose of the Study:

• To develop a hybrid intelligent framework for the early detection of liver disease.
• To integrate demographic segmentation with machine learning for improved diagnostic accuracy.

Main Methods:

• Utilized two datasets: Indian Liver Patient Dataset (ILPD) and a large-scale dataset.
• Stratified patients by age and gender into six groups for segment-specific model development.
• Evaluated 16 machine learning algorithms using feature selection, resampling, and hyperparameter optimization, integrating segment-specific models into a hybrid system.

Main Results:

• Achieved 94.2% accuracy on ILPD and 99.8% accuracy on the large dataset.
• Demonstrated consistent accuracy improvements across different demographic segments.
• Identified distinct biomarker importance based on age and gender, highlighting the need for tailored diagnostics.

Conclusions:

• The proposed framework offers a scalable, non-invasive tool for early liver disease detection.
• Combines demographic awareness, hybrid learning, and interpretability for personalized and accessible healthcare.
• Advances the clinical relevance and accessibility of liver disease diagnostics.