Radiomics and Artificial Intelligence Renal Cell Carcinoma

Radiomics and AI algorithms have demonstrated significant potential in improving the diagnosis and treatment of kidney cancer. ML-based kidney organ and tumor segmentation, such as the nnUNet algorithm from the Kits19 and Kits21 challenges, has achieved high Dice scores, with further research needed to assess its portability and accuracy in real-world settings. Additionally, these algorithms have shown success in classifying kidney masses into malignant and benign classes, predicting oncologic and perioperative outcomes, and calculating scoring systems such as RENAL, PADUA, and C-index. With continued improvement in areas such as renal cancer subtyping, TNM staging, pathologic tumor grading, and developing easily explainable algorithms, radiomics holds promise as a noninvasive and robust method for enhancing patient treatment of RCC.

Automated summary

Radiomics and AI algorithms have great potential in improving the diagnosis and treatment of kidney cancer, enabling tumor segmentation, classification, and prediction of treatment outcomes.