Purpose: To calculate pulse wave velocity (PWV) in the proximal pulmonary arteries (PAs) by cardiovascular magnetic resonance (CMR) using the transit-time method, and address respiratory variation, repeatability, and observer reproducibility.Materials and Methods: A 1.9-msec interleaved phase velocity sequence was repeated three times consecutively in 10 normal subjects. Pulse wave (PW) arrival times (ATs) were determined for the main and branch PAs. The PWV was calculated by dividing the path length traveled by the difference in ATs. Respiratory variation was considered by comparing acquisitions with and without respiratory gating.Results: For navigated data the mean PWVs for the left PA (LPA) and right PA (RPA) were 2.09 Ϯ 0.64 m/second and 2.33 Ϯ 0.44 m/second, respectively. For non-navigated data the mean PWVs for the LPA and RPA were 2.14 Ϯ 0.41 m/second and 2.31 Ϯ 0.49 m/second, respectively. No statistically significant difference was found between respiratory non-navigated data and navigated data. Repeated on-table measurements were consistent (LPA non-navigated P ϭ 0.95, RPA non-navigated P ϭ 0.91, LPA navigated P ϭ 0.96, RPA navigated P ϭ 0.51). The coefficients of variation (CVs) were 12.2% and 12.5% for intra-and interobserver assessments, respectively. Conclusion:One can measure PWV in the proximal PAs using transit-time in a reproducible manner without respiratory gating.

Cardiovascular magnetic resonance detects subclinical aortic atherosclerosis, can follow plaque burden over time, and confirms the presence of Glagov remodeling with preservation of the lumen despite progression of plaque. Cardiovascular magnetic resonance is well suited for the longitudinal follow-up of the general population with atherosclerosis, may help in the understanding of the natural history of atherosclerosis, and in particular may help determine factors to retard disease progression at an early stage.

Purpose: To improve three-dimensional (3D) volume-selective turbo spin-echo (TSE) carotid wall imaging by the addition of a novel body surface swallowing detection device. Material and Methods:A 3D volume-selective TSE sequence was used to image the carotid artery. A novel carbon-fiber motion device, positioned over the laryngeal prominence, was used to detect swallowing movement. An electrical output generated by coil movement was used to detect motion, and an algorithm was programmed to reject data acquired during swallowing and for a short period afterwards. Images were acquired with and without the algorithm and scored on a scale of 0 -5 by four independent blinded observers according to the clarity of the vessel wall, e.g., 0 ϭ poor image quality and 5 ϭ excellent quality images with little or no artifact. Results:The scans with the rejection algorithm on were scored higher than the scans without the algorithm. The comparison of scores with the algorithm on vs. the algorithm off were as follows: mean Ϯ standard deviation (SD) ϭ 3.76 Ϯ 0.25, 95% confidence interval (CI) ϭ 3.27-4.25 vs. 2.64 Ϯ 0.25, 95% CI ϭ 2.15-3.13; with good interobserver correlation (Kendall's W score 0.77). Conclusion:Image quality can be improved by the algorithm during acquisition. This can be achieved by a novel, anatomically positioned superficial device. This may help in prolonged 3D scans where a single movement can corrupt the entire acquisition.

No abstract

We welcome Laffon et al's interest and reply to their concerns about the validity of comparing their pulmonary artery (PA) pulse wave velocity (PWV) data (1) with our own (2).Given the scarcity of literature on the subject, we considered it reasonable to discuss their work [particularly as it employed cardiovascular magnetic resonance (CMR)]. Indeed, we have identified just two other non-invasive studies (3,4) that used the transit-time approach, like ours. Only one, which employed fluorocardiography, had comparable results [conus to right hilum, 2 m/sec (3) vs. 2.33 m/sec for right PA and 2.09 m/sec for left PA (2)].The transit-time method records the time a pulse takes to cover a known distance and provides a "weighted" measure of PWV over this path length. By using a high-temporal resolution acquisition to analyze flow arrival, ours was the first study to prove this to be achievable in the PAs with CMR. Conversely, Laffon et al approached PWV "locally" by estimating the elastic modulus using the equation proposed by Frank, and Bramwell and Hill (5). "Local" PWV has also been estimated by CMR in work reported after ours was submitted where the change in flow was related to the change in diameter in the early systolic phases of an MPA velocity map (6). In the 8 healthy volunteers included in the study, a mean PWV of 1.84 m/sec was found.It is our assertion that as long as the differences between the methods (as well as their limitations) are acknowledged clearly, discussion of all available measurements in this scarcely reported field will remain warranted [see the chapter on wave propagation in Milnor's Hemodynamics (5)].Laffon et al's main concern centers on the inclusion of the main PA's branch point in the examined region and the effects this might have on transit-time PWV. Because of the limitations imposed by pulmonary anatomy and temporal resolution, this was the only way to make the delay in pulse wave arrival discernible. Although future improvements in velocity mapping could limit study to a single vessel only, it will not stop reflections originating outside the examined region "contaminating" the area of interest.Although it was not the intention of this work to examine the issue of reflections or make comment on the behavior of the pulse wave at any specific position during its transit, we considered this problem when choosing our "marker" of pulse wave arrival. In normal subjects, where one could speculate that the backward wave would be both small and slow moving, we believed a point halfway up the wavefront would likely be relatively unaffected during its transmission and therefore robust. However, in pulmonary hypertension, where the wave profile changes to a far greater degree during propagation, the same point would likely be unrecognizable by the end of the path length. So, just as Fleischner et al (3) and Ring and Kurbatov (4) did over 50 years ago, we will be focusing on the wavefront foot, which we anticipate will prove least contaminated and hence easiest to study.In conclusion, we belie...