Quantification of aortic pulse wave velocity from a population based cohort: a fully automatic method

Area of Science:

• Cardiovascular Imaging
• Biomedical Engineering
• Medical Physics

Background:

• Aortic pulse wave velocity (PWV) quantifies aortic stiffness, a key predictor of cardiovascular events.
• Magnetic resonance imaging (MRI) offers a non-invasive method for PWV assessment.
• Accurate PWV calculation relies on aortic arch length and wave transit time.

Purpose of the Study:

• To develop and evaluate a fully automatic method for quantifying aortic pulse wave velocity (PWV) using MRI.
• To assess the performance of the automatic method in a large, multi-center population-based cohort.
• To establish the utility of automated PWV measurement in large-scale cardiovascular research.

Main Methods:

• Retrospective selection of 212 subjects from a multi-center heart-brain connection cohort.
• Acquisition of multi-slice 3D aortic scans and 2D velocity-encoded (VE) MRI scans.
• Development of multi-atlas segmentation for aortic arch length and algorithms for delineating aorta and deriving velocity-time flow curves from VE scans.

Main Results:

• The automatic method demonstrated high agreement with manual analysis for both aortic arch length and PWV.
• Mean absolute difference in aortic length was 3.3 ± 2.8 mm (p < 0.05) compared to manual measurements.
• Bland-Altman analysis showed minimal bias for PWV calculation across different methods (foot-to-foot, half-max, cross-correlation).

Conclusions:

• A fully automatic method for calculating aortic PWV from multi-center MRI data has been successfully developed and validated.
• The automatic method shows excellent agreement with manual analysis, making it suitable for large population studies.
• This automated approach significantly reduces the manpower required for PWV assessment in extensive research cohorts.