Predictive Modeling for Diagnosing Infectious Pleuritis in Cats Using XGBoost: A Comparative Study

Infectious pleuritis is a life-threatening condition in cats, characterized by the accumulation of inflammatory fluids in the pleural cavity. Accurate and timely diagnosis is crucial for effective treatment. This study presents a machine learning approach to predict infectious pleuritis using clinical data collected from 181 feline cases between 2019 and 2021. Six models were evaluated: Logistic Regression, Random Forest, LightGBM, XGBoost, Support Vector Machine (SVM), and CatBoost. Among these, XGBoost demonstrated the best performance, achieving an accuracy of $88 \%$ by effectively addressing class imbalance and complex feature interactions. This paper provides insights into the methodology, model comparisons, and clinical implications of using predictive analytics for diagnosing feline infectious pleuritis.