Routine cine-CMR segmentation via a novel automated algorithm (LV-METRIC) for assessment of aortic physiology: a clinical validation study
Journal of Cardiovascular Magnetic Resonance volume 17, Article number: P387 (2015)
Routine cine-CMR is widely used to assess cardiac structure and function. Partial voxel interpolation has been shown to yield improved agreement with phantom derived chamber volumes and necropsy evidenced LV mass; the utility of partial voxel interpolation for assessment of aortic physiology has never before been tested.
Cine-CMR (SSFP) was performed on 1.5 Tesla (GE) scanners; pulse sequence parameters were equivalent to those for routine CMR (typical TR 3.4 msec, TE 1.14 msec, flip angle 60ο, temporal resolution 30 msec). Images were acquired in conventional cardiac (2, 3, 4 chamber) long axis or axial imaging planes. Aortic area was uniformly measured in a non-aneurysmal location within the mid-descending thoracic aorta: Cine-CMR was quantified via a novel "partial voxel" segmentation algorithm (LV-METRIC) that accounts for relative proportion of blood within each individual imaging voxel. Maximum (systolic) and minimum (diastolic) aortic areas and brachial pulse pressure were used to calculate distensibility, a measure of arterial compliance, of the mid-descending thoracic aorta.
32 subjects were studied, among whom 22 had genetically-mediated aortopathies (13 bicuspid aortic valve [BAV], 9 Marfan syndrome [MFS]) and 10 were normative controls. Aortopathy subjects were similar to controls in age, gender, pulse pressure and body size (all p=NS). Aortic indices, compared between MFS, BAV and control groups are shown in Figure 1. As shown, absolute aortic size (measured in a non-aneurysmal region) was similar between groups (all p=NS). Dynamic change in aortic area (Δ Area) was lesser among MFS vs. controls (p=0.004) as well as BAV (p=0.03), but not between BAV vs controls (p=0.99). Aortic distensibility, as measured in all aortopathy subjects and 6 normative controls, demonstrated lower values among MFS subjects as compared to normative controls (p=0.007), with a similar trend when MFS and BAV groups were compared (p=0.08). There was no significant difference in distensibility between BAV vs controls (p=0.43).
Routine cine-CMR can discern altered aortic physiology in non-aneurysmal regions in subjects with MFS. Larger longitudinal studies are needed to further evaluate the prognostic utility of cine-CMR segmentation, including use of central aortic blood pressure, as a potential biomarker of early aortic disease.
About this article
Cite this article
Singh, P., Codella, N.C., Wang, Y. et al. Routine cine-CMR segmentation via a novel automated algorithm (LV-METRIC) for assessment of aortic physiology: a clinical validation study. J Cardiovasc Magn Reson 17 (Suppl 1), P387 (2015). https://doi.org/10.1186/1532-429X-17-S1-P387