This study aims to develop and validate a contrast-enhanced CT-based deep-learning model for automatic and accurate preoperative discrimination between T1-T2 and T3 renal cell carcinoma. By quantifying the model's diagnostic performance on an independent test set-using AUC, sensitivity, specificity, positive/negative predictive values, and decision-curve analysis-we will establish a decision-support tool that can be seamlessly integrated into clinical PACS, thereby reducing staging errors, refining surgical planning, and improving patient outcomes.

Inclusion Criteria:

1. Histopathologically confirmed renal cell carcinoma on postoperative specimen.
2. Preoperative contrast-enhanced CT performed at our institution with slice thickness ≤ 1 mm and complete DICOM datasets.
3. Postoperative pathologic staging clearly defined as pT1a-T2b or pT3a.
4. CT image quality deemed adequate for analysis.

Exclusion Criteria:

• 1. Pathologic subtype other than RCC. 2. Images with severe artifacts.

Patients who underwent surgery at the Department of Urology, Peking University First Hospital, with postoperative pathological confirmation of renal cell carcinoma (RCC), and who also have complete preoperative contrast-enhanced CT datasets (slice thickness ≤1 mm, lossless DICOM) and definitive pathological staging of pT1a-T2b or pT3a.