MR Diffusion Models for Non-invasively Assessing Renal Interstitial Fibrosis and Annual Renal Function Loss in Chronic Kidney Disease

Motivation: Non-invasive and reliable biomarkers for grading interstitial fibrosis (IF) and predicting progression of chronic kidney disease (CKD) are needed. Goal(s): To explore the effectiveness of seven diffusion models (DWI, IVIM, DKI, CTRW, FROC, SEM, and sADC) in assessing renal IF and annual renal function loss in CKD. Approach: The performances of diffusion models in distinguishing between mild and moderate-to-severe IF were evaluated. The association between diffusion metrics and estimated glomerular filtration rate (eGFR) slopes were investigated. Results: Combined model of cortical fIVIM, MKDKI and eGFR outperformed individual clinical model in grading IF, with cortical fIVIM and αSEM as potential biomarkers for CKD progression. Impact: This study identifies the optimal diffusion metrics for grading renal IF and predicting short-term progression of CKD, aiding in the guidance of personalized management and the prevention of poor outcomes.