Three‐dimensional CT angiography: A new technique for imaging microvascular anatomy

Abstract

To the Editor, Clinical Anatomy:We would like to congratulate Tregaskiss et al. for theirstudy utilizing computerized tomography angiography(CTA) for mapping the vasculature of the body wall(Tregaskiss et al., 2007). This is a significant contribution tothe literature in this evolving field of anatomical research. Asmentioned in their article, CTA has only recently been sug-gested as an imaging modality for mapping vasculature inanatomical research, and even more recently as a modalityfor preoperative imaging in microsurgical, reconstructiveflap surgery. We have conducted similar studies, in the set-tings of both cadaveric research and clinical studies (in theprocess of publication in this journal (Rozen et al., in press-b)), utilizing plain radiography, ultrasonography, CTA, andmagnetic resonance angiography (MRA), the findings ofwhich complement those of Tregaskiss et al.Tregaskiss et al. discuss in their introduction the motiva-tion for three-dimensional (3D) CTA being the differentiationof vessels between tissue layers and their course betweenlayers. Our early anatomical research aimed to address thiswith cross-sectional studies. Cross-sectional studies canaccurately map the course of vessels in a similar fashion toCTA, as the cross sections permit analysis of the third spatialdimension (depth) (Rozen et al., in press-a). However,cross-sectional studies are significantly limited by time,labor-intensiveness, and inadequate spatial resolution.We also undertook CTA studies of our cadaveric speci-mens, with 10 cadaveric hemiabdominal walls injected, har-vested, and scanned in a similar fashion to Tregaskiss et al.Our CT scanner did differ, in being a 64 multidetector row CTscanner (Siemens Somatom Sensation Cardiac 64). Ourreformatting also differed, in that we created maximum in-tensity projection images in addition to 3D volume-renderedtechnique images. These changes in scanning and reformat-ting technique, we believe, enable a greater spatial apprecia-tion of the vasculature, and achieve the aim of high resolu-tion images of the course of vessels relative to tissue planes.Tregaskiss et al. have been diligent in performing an eval-uation of the efficacy of CTA by measuring the diameters ofvessels seen on CTA. This was an effective method, but doesnot achieve a formal evaluation of the accuracy of the CTAtechnique. We have performed a double-blinded anatomicdissection, evaluating for all musculocutaneous vessels ineach specimen, after a CTA of the same specimen had beenperformed (see Figs. 1 and 2). A vessel map was created foreach technique and the accuracy of CTA compared to ana-tomic dissection (total number on CTA ¼ 154 vessels). Wefound that the mapping of vessels was highly accurate, withonly eight false positives and six false negatives out of 154vessels seen on CTA, establishing a sensitivity of 96% anda specificity of 95% for the mapping of such vessels. Ourresults confirm the successful use of CTA for vessel mappingin cadaveric studies, and as a potential tool for preoperativeimaging prior to reconstructive flap surgery.Our research has demonstrated that CTA can be used witha high accuracy in anatomical studies such as that of Tregas-kiss et al, and suggests that this may be an ongoing modalityfor imaging vasculature in both anatomical studies and forpreoperative imaging. Our preliminary clinical studies haveshown high concordance of CTA with operative findings (seeFig. 3), and demonstrated high sensitivities (up to 95%) andpositive predictive values (up to 95%) to match the accuracyof the anatomical studies (Rozen et al., in press-c).Tregaskiss et al. also briefly mention the potential use ofMRA for preoperative imaging of these blood vessels, butacknowledge the lack of research in this area. Of note, wehave performed clinical studies of MRA for mapping theabdominal wall vasculature prior to breast reconstructionflaps, and have had good preliminary results (Rozen et al.,in press-d). As demonstrated in Figure 4, contrast-enhanced vessels are able to be clearly identified within thesurrounding soft tissues, but are not characterized with thesame resolution of CTA.The study by Tregaskiss et al. is an important study ofthe utilization of advancing imaging technologies in basicanatomical research, and will certainly form the foundationfor clinical application in the field of reconstructive surgery.Although this type of anatomic cadaveric study does notnecessarily reflect the accuracy of CTA in clinical settings,our studies certainly complement this work in delineatingthe clinical role for CTA in preoperative localization of themusculocutaneous perforating vessels of the abdominalwall for planning reconstructive surgery with the abdominalwall integument.