The use of central venous catheters has increaseddramatically over the past 40 years [1]. These devices areessential for many clinical treatments including access forchemotherapy, antibiotics, and blood transfusions [2].Indwelling catheters terminating in a port are particularlyuseful for intermittent treatment for malignancies [3]. Ad-vantages include protection of the device from damage orinfection by the overlying skin and a reduction in veni-puncture injuries and infusion complications [1,4,5]. In thisstudy, we investigated the Angiodynamics Vortex Smart PortMini (Angiodynamic Inc, Manchester, GA). These portswere inserted into the upper arm with the tip of the attachedcatheter at the right atrial/superior vena cava junction. Thisparticular arm port model is approved for power injection ofcomputed tomography (CT) contrast agent, which hastraditionally not been possible with other arm ports. Becausearm ports are not typically used for contrast injection, theobjective of this study was to evaluate the image quality ofCTs obtained via arm port power injection compared withcontrast injection of a peripheral vein.MethodsWe performed a retrospective cohort study of patientswho had a power injectable arm port inserted between July2013 and June 2014. Ethics approval for this study was ob-tained from our local biomedical research committee(University of Saskatchewan) and data review was incompliance with provincial and national privacy legislation.Subjectswereidentifiedbyreviewingadatabaseofpatientswhohadundergonecontrast-enhancedCT(n¼54).Toreducevariability from timing of contrast bolus, we analysed onlyscans with a mixed phase of contrast (contrast present withinthesystemicarterialandportalvenoussystems).Theinclusioncriteria were patients who had a combined chest-abdomen-pelvis CT scan via an arm port injection, and who also had aprior CT scan of the same type with injectionvia a peripheralarm vein. Therefore, each subject served as his or her owncontrol. Eighteen subjects met these inclusion criteria.CT images were reviewed on Philips iSite RadiologyPicture Archive and Communication System (KoninklijkePhilips Electronics, Best, The Netherlands). Scan dates rangedbetween July 2012 and July 2014 with CT slice thicknessranging from 3.8-5 mm and volume CT dose index rangingfrom 2.2-20.43 mGy. CT scans were from 4 acute care centres(Royal University Hospital, Saskatoon City Hospital, CypressRegional Hospital, and Battlefords Union Hospital).Assessment of objective image quality was performedby measuring signal-to-noise (SNR) and contrast-to-noise(CNR) ratios, as described by Heyer et al [6]. SNR andCNR were determined at 3 anatomical levels: 1) aortic arch;2) main pulmonary artery (MPA) and right pulmonary artery(RPA); and 3) descending aorta and left atrium (Figure 1).The following equations were used for each structure, whereSI