Renal Cell Carcinoma (RCC) Staging Using Neural Networks: A Deep Learning Approach

The most prevalent kidney cancer known as Renal Cell Carcinoma (RCC) needs precise staging information to plan effective treatments properly. The traditional staging evaluation method depends on human assessment of tumors but such measurements lack accuracy because they involve subjective methods. The proposed system implements YOLOv8 for tumor segmentation together with MobileNetV2 classification to build an automated RCC staging process. We process nephrographic CT scan pictures due to their enhanced tumor visibility. The model demonstrates excellent accuracy when identifying T1a T1b T2 and T3 stages of tumors by achieving 89.68% average accuracy coupled with 0.94 precision and 0.837 recall and 0.89 F1-score. The paper presents complete results that show loss evolution through iterations as well as accuracy metrics alongside confusion matrices for testing and validation data sets. The paper contains a thorough explanation of MobileNetV2 architecture together with proper citation for all image and external written references. The research furnishes a complete protocol for RCC staging that applies neural networks in an exact and replicable manner.