Clinical validation of quantitative carotid MRA

Abstract

Objectives: To determine the accuracy of a new software for automated quantitative 3D magnetic resonance angiography (MRA) analysis and to validate the findings with corresponding measurement on maximal intensity projection (MIP) images and digital subtraction angiography (DSA) images. Furthermore, to determine the inter-observer variability on carotid artery measurements using the three different methods. Methods: A retrospective review of arch angiograms and carotid first-pass, gadolinium-enhanced 3D-MRA, was performed on 10 patients. Common carotid arteries were measured by two individual investigators using DSA and MRA images. Inter-observer variability and the correlation between DSA, automated MRA evaluation and evaluation on MIP images was determined. Results: Inter-observer agreement on common carotid artery measurements was excellent for DSA and automated MRA evaluation. For each observer, comparison between DSA and automated MRA evaluation showed no statistically significant difference, whereas on MIP measurements, greater discrepancies were shown. Conclusions: Automated MRA evaluation allows reproducible measurements of carotid arteries between different observers. It can be used as an accurate method for analyzing vessel dimensions based on MRA images.