Value of multi-parametric diffusion-weighted imaging in pathological grading of clear cell renal cell carcinoma

Motivation: Accurate preoperative grading of clear cell renal cell carcinoma (ccRCC) is essential for treatment decisions, particularly in patients with comorbidities. Goal(s): This study aims to investigate the utility of various diffusion-weighted imaging (DWI) models, including mono-exponential, intravoxel incoherent motion (IVIM), diffusion kurtosis imaging (DKI), and continuous time random walk (CTRW), in preoperative ccRCC grading to facilitate personalized treatment strategies. Approach: We conducted MRI scans on 105 ccRCC patients with multi-b value DWI and employed machine learning for model construction for ccRCC grading. Results: This study demonstrates the protential of multi-parametric DWI models for accurate ccRCC grading, yield an AUC of 0.861. Impact: This study has the potential to reshape the landscape of preoperative ccRCC grading, promote objectivity in treatment planning.