AI-Based Aortic Calcification Score with 4D Flow MRI to Evaluate Aortic Remodeling and Predict Cardiovascular Events in Chronic Kidney Disease

Motivation: Traditional imaging techniques are limited in assessing both aortic structure and function, especially chronic kidney disease (CKD) patients. Goal(s): This study aims to explore the clinical value of an AI aortic calcification score (AI-ACS) and 4D Flow MRI to reveal aortic remodeling characteristics in CKD patients. Approach: 50 CKD patients and 45 healthy controls underwent 4D Flow MRI and chest CT scans. Results: The combined Clinical + AI-ACS + 4D Flow model outperformed several individual models, achieving an AUC of 0.92 in the training cohort and 0.67 in the testing cohort for predicting major adverse cardiovascular events (MACE). Impact: The combination of AI-ACS and 4D Flow MRI provides a quantitative assessment of aortic remodeling in CKD patients, enhancing cardiovascular risk prediction and offering a novel approach for cardiovascular management in CKD.