Performance of Machine Learning Models Based on Medical Imaging in Predicting Pathological Grade of Clear Cell Renal Cell Carcinoma

Cancer research · SCLC

Predicting clear cell renal cell carcinoma (ccRCC) pathological grade preoperatively is critical for clinical management.

This systematic review and meta‑analysis evaluated the diagnostic accuracy of radiomics and deep learning models for pre‑operative prediction of high‑grade clear cell renal cell carcinoma (ccRCC). 43 studies comprising 12,675 patients were pooled, yielding an area under the summary receiver operating characteristic curve of 0.89 with pooled sensitivity 0.79 and specificity 0.85. Decision‑curve analysis suggested a net clinical benefit over conventional strategies across a wide threshold range, and Fagan nomogram analysis indicated that a positive test increased the post‑test probability of high‑grade disease to 70 % whereas a negative test reduced it to 10 %. The study highlights the potential of machine learning‑based imaging to inform surgical decision‑making and underscores the need for external validation to confirm performance in independent cohorts.

Accurate pre‑operative grading of ccRCC informs surgical planning (e.g., radical versus partial nephrectomy) and prognostication. Non‑invasive, radiomics‑ or deep‑learning‑based approaches could reduce reliance on percutaneous biopsy, mitigate procedural risks, and streamline clinical workflows, ultimately improving patient outcomes.

Evidence level: Számítógépes vagy elméleti. Modellből vagy adatbányászatból származó jel.

Original source

Related signals