Measurements After Image Post-processing Are More Precise in the Morphometric Assessment of Thoracic Aortic Aneurysms: An Intermodal and Intra-observer Evaluation.

OBJECTIVES Precise pre-procedural anatomical analysis of aneurysmal anatomy is essential for successful thoracic endovascular aortic repair (TEVAR). Since surgeons and radiologists have to perform multiple measurements in the same patient, high intra-observer reliability of any imaging method is mandatory. Commercially available three dimensional (3D) post-processing techniques are expected to be superior to conventional two dimensional multiplanar reconstructions (MPRs) derived from computed tomography angiograms (CTAs). However, few data exist to support this view. This study aims to evaluate the intermodal and intra-observer differences using 3D software (3surgery) in descending thoracic aortic aneurysms (dTAAs). METHODS Pre-operative CTAs (performed between 2004 and 2010) of 30 dTAAs (mean maximum diameter 61.4 ± 13 mm) were assessed by three independent investigators with different experience in the measurement of aortic pathologies. Intra-observer reliability and intermodal differences (3D vs. 2D) were investigated using pre-specified measurement points (distances of total length, maximum diameter, proximal and distal landing zones). Statistical analyses were performed using the Bland-Altman method and a mixed regression model. RESULTS Intermodal comparison showed that 2D measurements significantly underestimate the measured distances (maximum diameter 3.7 mm [95% CI -5.3 to -2.1] and landing zone maximum 1.4 mm [95% CI -2.0 to -0.2] shorter with 2D, p < .05). In almost all 3D measurements, all investigators showed lower variability comparing the intra-observer differences, most notably in the measuring point total length (reduction of the SD up to 7.9 mm). CONCLUSIONS These data show that both techniques led to significant measurement disparity. This occurs especially at the point of indication (maximum diameter) and the total length of the aneurysm (important for correct stent graft selection). But overall the variability is reduced with the 3D technique, which also tends to measure greater distances. The use of post-processing software therefore leads to more precise device selection for TEVAR in TAA.