Determination of Malignancy Risk Factors Using Gallstone Data and Comparing Machine Learning Methods to Predict...

Area of Science:

• Digestive system disorders
• Oncology
• Computational biology

Background:

• Gallstone disease is a common condition with multifactorial risk factors.
• Some risk factors for gallstones may also increase cancer risk.
• Understanding these links is crucial for early detection and prevention.

Purpose of the Study:

• To investigate the association between gallstone disease and malignancy.
• To apply advanced machine learning (ML) algorithms for this analysis.
• To identify significant predictors of malignancy in patients with gallstones.

Main Methods:

• Analysis of a dataset of approximately 1000 patients.
• Utilized six ML algorithms: random forests, SVMs, MLP, MLP_PT, naive Bayes, and TabPFN.
• Performance evaluated using metrics like AUC, accuracy, sensitivity, and specificity.

Main Results:

• Age, body mass index (BMI), and hormone replacement therapy (HRT) history were identified as key predictors of malignancy.
• The Tabular Prior-data Fitted Network (TabPFN) model demonstrated superior performance.
• TabPFN achieved high accuracy and effectiveness in predicting malignancy.

Conclusions:

• Cutting-edge ML methods can uncover complex clinical data relationships.
• TabPFN shows significant promise for predictive modeling in gallstone-related malignancy.
• Findings offer insights for personalized patient management and risk assessment.