Thorac Cardiovasc Surg 2015; 63 - OP155
DOI: 10.1055/s-0035-1544407

New Automatic Modality for Multislice Computed Tomography Prediction of Optimal Implantation Angle for Transcatheter Aortic Valve Implantation

A. Van Linden 1, J. Blumenstein 2, H. Möllmann 2, W.-K. Kim 1, 2, T. Ziegelhöffer 1, S. Alkaya 1, J. Kempfert 1, T. Walther 1
  • 1Kerckhoff-Klinik, Herzchirurgie, Bad Nauheim, Germany
  • 2Kerckhoff-Klinik, Kardiologie, Bad Nauheim, Germany

Objectives: Different imaging techniques, including intraprocedural rotational angiography (DynaCT, Siemens Healthcare), have been suggested for predicting the optimal perpendicular C-arm angulation for transcatheter aortic valve implantation (T-AVI). The prototype software Aortic ValveGuide (AVG) of the DynaCT workstation also allows for automatic processing of conventional multi slice computed tomography (MSCT) images. Aim of this study was to analyze predicted C-arm angulation using AVG in the established DynaCT compared with AVG in preoperative MSCT.

Methods: 84 image-datasets of patients with comparable patient-position during MSCT and T-AVI were included (36% transapical). DynaCT and MSCT raw-datasets were analyzed using AVG software prototype with automatic detection of the relevant landmarks of the aortic root. Two different perpendicular angulations were determined and compared. One angulation showed the left main on the right side of the screen (Angle1) and the other angulation had the left/right commissure in the middle (Angle2).

Results: Automatic landmark detection worked well in all image-datasets. Minor manual adjustments were necessary in 22 patients. Mean Angle1 for DynaCT was LAO/CRA 29 ± 17/20 ± 15° and for MSCT 29 ± 17/17 ± 13° (p < 0.05). Mean Angle2 for DynaCT was LAO/CRA 41 ± 10/31 ± 11° and for MSCT 41 ± 10/31 ± 9° (p = 0.48). Mean deviation independent from CRA/CAU direction was 7° with a significant correlation between the two methods (r = 0.7 and 0.6, respectively, p < 0.001).

Conclusions: Aortic ValveGuide assessment of preoperative MSCT-scans seems to equally predict perpendicular C-arm angulation for T-AVI compared with intraprocedural DynaCT. Automatic processing of the image-datasets reduces the effort for MSCT analysis and may lead to more reproducible prediction. The accuracy seems to be patient-dependent rather than method-dependent.

Fig. 1 Similar angulation in MSCT (left) and DynaCT (right).