Detection of Renal Cell Carcinoma Based on Clinical Data Using Machine Learning Approach

Renal cell carcinoma (RCC) is a highly hetero-geneous malignancy. We propose a machine learning (ML)-based predictive model utilizing high-dimensional clinical data to enhance RCC classification. To address high dimensionality and small sample challenges, we implement principal component analysis (PCA) as feature extraction technique. We evaluate multiple models, including Random Forests (RF), XGBoost, LightGBM and CatBoost, Support Vector Machine (SVM) and a voting ensemble model. These models are used to classify RCC patients. The result from predictive analysis suggests that the ML model can classify malignant patients from benign patients. Since early-stage detection is often limited by imaging techniques, ML-based detection could play a significant role in improving RCC diagnosis.