Using Deep Learning on Contrast-Enhanced CT to Differentiate Early and Advanced Stages of Renal Cell Carcinoma

What this Trial Is About

Researchers are developing and validating a deep-learning model based on contrast-enhanced CT scans to automatically distinguish between early-stage T1-T2 and more advanced-stage T3 renal cell carcinoma before surgery. This observational study aims to measure the accuracy of the model using various statistical tools including AUC, sensitivity, specificity, and predictive values. The goal is to create a decision-support tool that can be integrated into clinical imaging systems to reduce staging errors and improve surgical planning and patient outcomes. The study involves analyzing preoperative contrast-enhanced CT images collected retrospectively, without any intervention or treatment being administered. The CT scans used have a slice thickness of 1 mm or less and must meet quality standards suitable for analysis. Participants postoperative specimens confirm the renal cell carcinoma diagnosis and staging, which serves as a reference for evaluating the models diagnostic performance. Participants will have undergone preoperative CT imaging and surgery, with their data analyzed retrospectively during the study period from 2024 to 2027. Researchers will assess the diagnostic accuracy of the deep-learning model compared to the pathologic staging results. The study includes healthy volunteers and covers an age range from 18 to 85 years. There is no active treatment or follow-up involvement required from participants during the study.

Deep Learning for Automated Discrimination Between Stage T1-T2 and T3 Renal Cell Carcinoma on Contrast-Enhanced CT