Phantom investigation on the accuracy of different pulse sequences for the determination of arterial distensibility

Introduction Arterial stiffness has long been recognized to play a central role in the development of cardiovascular (CV) disease. A significant relationship has been demonstrated between reduced aortic compliance and fatal and non-fatal CV events independently of classical CV risk factors [1]. Accurate measurement of aortic distensibility in vivo would constitute a powerful tool for monitoring patients at high CV risk and testing plaque-reducing pharmaceuticals and therapies. As the cross-sectional distensibility coefficient is defined as the relative change in luminal area divided by the corresponding pressure change over the cardiac cycle (D = ∆A/(Adiast·∆P)), the relative area variation curve over time needs to be determined. Despite systematic and random errors have been shown to affect 2D phasecontrast (PC) images due to intravoxel phase dispersion, partial volume effects and misalignment of flow-encoding gradients and flow direction [2], a number of investigations [3-5] have used this technique to determine arterial distensibility. New approaches to the post-processing of PC images have been proposed to overcome these limitations [6-7] while other MR techniques have been explored to extract the area variation curve over the cardiac cycle [8]. The aim of this work was to assess the accuracy of different MR pulse sequences for the determination of the luminal area variation curve over time by means of in vitro experiments conducted on a human-tissue-mimicking phantom and to make direct comparisons with high resolution digital photography (HRDP) as the defined goldstandard.