Evaluation of a new imaging software for aortic endograft planning

The aim of this study was to evaluate a new 3D Workstation workflow (EVAR Assist, Advantage Windows, GE Healthcare, Chalfont, UK) (EA-AW) designed to simplify complex EVAR planning.

All pre-operative computed tomography (CT) scans of patients who underwent repair at our institution of a complex aortic aneurysm using fenestrated endovascular repair (f-EVAR) between January and September 2014, were reviewed. For each patient, imaging analysis (12 measures: aortic diameters and length and "clock position" of visceral artery) was performed on two different workstations: Aquarius (TeraRecon, San Mateo, CA, USA) and EA-AW. According to a standardized protocol, three endovascular surgeons experienced in aortic endograft planning, performed image analyses and data collection independently. We analyzed an internal assessment between observers (on the Aquarius 3DWS) and an external assessment comparing these results with the planning center (PC) data used to custom the fenestrated endograft of the patients enrolled in this study. Finally, we compared both 3DWS data to determine the accuracy and the reproducibility. A p-value < .05 was considered as statistically significant. Complete agreement between operators was defined as 1.0.

Intra- and inter-observer variability (interclass correlation coefficients - ICC: 0.81-.091) was very low and confirmed the reliability of our planners. The ICC comparison between EA-AW and Aquarius was excellent (> 0.8 for both), thus confirming the reproducibility and reliability of the new EA-AW application. Aortic and iliac necks diameters and lengths were similarly reported with both workstations. In our study, the mean difference in distance and orientation evaluation of target vessels evaluated by the two workstations was marginal and has no impact on clinical practice in term of device manufacturing.

We showed that complex EVAR planning can be performed with this new dedicated 3D workstation workflow with a good reproducibility.