Dual-Source Dual-Energy CT Angiography of the Supra-Aortic Arteries with Tin Filter: Impact of Tube Voltage Selection

Abstract

Automatic bone and plaque subtraction (BPS) in computed tomographic angiographic (CTA) examinations using dual-energy CT (DECT) remains challenging because of beam-hardening artifacts in the shoulder region and close proximity of the internal carotid artery to the base of the skull. The selection of the tube voltage combination in dual-source CT influences the spectral separation and the susceptibility for artifacts. The purpose of this study was to assess which tube voltage combination leads to an optimal image quality of head and neck DECT angiograms after bone subtraction. Fifty-one patients received tin-filter-enhanced DECT angiograms of the supra-aortic arteries using two voltage protocols: 24 patients were studied using 80/Sn140 kV and 27 using a 100/Sn140 kV protocol, both protocols with an additional tin filter. A commercially available DE-CTA BPS algorithm was used. Artificial vessel erosions in BPS maximum intensity projections (four-level Likert scale with CTA source data as reference) and vessel signal-to-noise ratio (SNR) were assessed in the level of the shoulders and the base of the skull in each patient and compared. At the level of the shoulder, 100/Sn140 kV achieved higher SNR (23.4 ± 6.4 at 80/Sn140 kV vs. 35.1 ± 11.8 at 100/Sn140 kV; P < .0001) with less erosions (erosion score 3.9 ± 0.4 in 80/Sn140 kV vs. 2.1 ± 1.3 in 100/Sn140 kV; P < .0001) than 80/Sn140 kV. At the level of the skull base, erosion scores and objective image quality of arterial segments were comparable with both protocols (P = .14). The 100/Sn140 kV protocol achieved more favorable results for BPS of the supra-aortic arteries than the 80/Sn140 kV protocol.