Appearance of a curvilinear interface in the ascending aorta which simulates an aortic dissection has been reported when using short-scan-time acquisitions (1 s) and not when using ultrafast CT (50 ms scan time) [1][2][3][4][5][6]. The artifact has been observed on both conventional and spiral scans. This artifact is thought to be related to the motion of the aortic wall in the interval time from end diastole to end systole [1, 5, 6]. Recently, it has been suggested that segmented images from data which have been collected during 0.6 s of a full 1-s scan of a conventional axial image can be retrospectively reconstructed to eliminate this artifact [1]. Spiral CT, which has gained widespread acceptance, requires interpolation to synthesize transaxial images from the volume data set [7]. The simplest approach is linear interpolation between spiral projection data sets from adjacent turns (i. e. 360°apart). However, using this interpolation volume, averaging artifacts due to the broadening of the section sensitivity profile have been noted [8, 9]. Consequently, all current spiral CT scanners reorder the projection data and perform interpolation from views separated by 180°(180°linear interpolation) [9]. In our previous personal experience (unpublished data), we often observed motion artifacts in the proximal part of the ascending aorta which simulate aortic dissection, using a 360°linear interpolation algorithm, as well as other series [3,4]. In several cases these artifacts disappeared when a 180°linear interpolation algorithm was used (Fig. 1).The goal of this study was to evaluate the prevalence of the pseudo-aortic dissection using a 180°linear interpolation algorithm. To our knowledge, the prevalence of the pseudo-aortic dissection using a 180°linear interpolation algorithm, in a large series, has not been reported.
Materials and methodsA total of 100 consecutive dynamic contrast-enhanced spiral chest CT examinations, taken during a 1-month interval, performed on a Somatom Plus S CT scanner (Siemens Medical Systems, Erlangen, Germany) for a variety of non-selected indications, excluding clinical suspicion of aortic dissection (i. e. lymphoma followup, lung-cancer staging, mediastinal lymph nodes), were obtained for assessment of the thoracic aorta. The patients included 41 women and 59 men, 19-85 years old (mean age 57 years). Spiral chest CT examinations were performed during a single breath-hold using 10 mm/s table motion, 10 mm collimation, and 10 mm contiguous reconstructions from the lung apex to the diaphragm. The contrast enhancement was provided by a monophasic bolus IV infusion of 120-150 ml of contrast material injected at a rate of 2 ml/s via an antecubital vein. Scanning began 30 s after the initialization of the bolus. Contiguous reconstruction were obtained with a 180°linear interpolation algorithm.The CT examinations were independently and retrospectively reviewed by two experienced radiologists in vascular radiology (P. L. and F. G.). Studies were interpreted as positive if at least one image sho...