Definition of the aortic valve plane by means of a novel dedicated software program: Proof of concept and validation with multi slice computed tomography

Abstract

Background: Proper selection of the C-arm angulation during transcatheter aortic valve implantation (TAVI) is essential. Currently multi slice computed tomography (MSCT) is increasingly used to determine the C-arm angulation and is considered the gold standard. Yet, MSCT is not available in the catheterization laboratory and requires expertise. We sought to evaluate the accuracy of a novel software providing the optimal angiographic projection (i.e. three cusps on one line, equally in size and posterior, right and left cusp sequence- optimal angiographic projection [OAP]) based on two contrast angiograms.

Methods: Software was validated by comparing the difference in OAP between MSCT (reference) and software (CAAS A-Valve) in 67 patients and by qualitative comparison of the aortograms performed in accordance to MSCT and CAAS A-Valve defined OAP in 20 patients by 2 blinded assessors.

Results: The median (IQR) difference in the OAP between MSCT and software was 12.4 (7.5-20.1) and 18.9 (10.8-28.5) degrees. For the qualitative comparison, an overall score ≥ 2 (acceptable-perfect) of 90% was reached by both observers for OAP derived by MSCT. It was 85% and 90% (observer 1 and 2) for OAP derived from the software and was 90% after consensus. The interobserver variability (mean difference and SD) for assessment of OAP curve and OAP itself was 1.6 (SD = 5.8, p = .23) and 9.5 (SD = 18.4) degrees (p = .03) respectively.

Conclusions: The CAAS A-Valve software adequately defines the OAP and is clinically useful in patients who do not have a MSCT prior to TAVI.