Dual-Energy Spectral Computed Tomography With Adaptive Statistical Iterative Reconstruction for Improving Image Quality of Portal Venography.

Abstract

The aim of the study was to investigate whether spectral computed tomography (CT) plus adaptive statistical iterative reconstruction (ASIR) could improve imaging quality of computed tomography portal venography (CTPV). Sixty-four patients underwent CTPV, with CT number of portal vein (PV) less than 150 HU in portal phase, were divided into 2 groups A (n = 31), using standard 120-kVp protocol. group B (n = 33), using spectral CT protocol. Standard 120-kVp images were reconstructed with 40% ASIR, and monochromatic images at 60 keV were reconstructed with the following 4 ASIR percentages: 0%, 20%, 40%, and 60%. The CT number, image noise, and contrast-to-noise ratio in main PV were measured. The maximum intensity projection and volume-rendering images were used for subjective evaluation. These 2 kinds of results were statistically analyzed. The contrast-to-noise ratio and subjective scoring of PV increased gradually from 120-kVp images to 60% weight ASIR (3.44 ± 0.95, 4.58 ± 1. 59, 5.26 ± 1.85, 6.18 ± 2.18, and 7.39 ± 2.65 and 4.35 ± 1.17, 6.21 ± 1.29, 6.48 ± 1.35, 6.85 ± 1.28, and 7.00 ± 1.19). There were statistically different for the 5 groups (P < 0.001). The CT number of the PV in the 60-kiloelectron volt spectral images had higher than the 120-kVp images (P < 0.001). The noise of 120 kVp was significantly higher than those of 60% ASIR and significantly lower than those of 0% ASIR (both P < 0.001), and there were no significant differences between 120-kVp, 20% ASIR, and 40% ASIR (P = 0.107 and 1.000, respectively). The diagnostic acceptability was highest at 40% ASIR. Forty-percent ASIR addition to the 60-kiloelectron volt monochromatic image could improve image quality of CTPV comparing with conventional 120-kVp images.