Automated post-processing of longitudinal 4D flow MRI in patients with an abdominal aortic aneurysm

Motivation: 4D flow MRI-derived parameters may enhance rupture risk prediction in abdominal aortic aneurysm (AAA) patients. Goal(s): We aimed to develop an automatic post-processing pipeline for 4D flow MRI in AAA patients, incorporating robust segmentation and aneurysm identification, to assess longitudinal 4D flow data and enable parameter validation. Approach: Included AAA patients underwent two MRI-scans with a half-year interval. An nnU-Net was trained to segment the aorta from 3D Cine bSSFP. Results: The automated post-processing pipeline extracted the aneurysm and enabled aneurysm specific parameter analyses. An inverse trend was observed between change in wall shear stress and change in diameter over the study period. Impact: We developed an automated pipeline for improved segmentation, aneurysm extraction and 4D flow MRI parameter calculation in patients with abdominal aortic aneurysms and provided initial results based on a longitudinal dataset. Ultimately, these insights may enhance rupture prediction.