Feasibility of virtual nonenhanced images derived from single-source fast kVp-switching dual-energy CT in evaluating gastric tumors

Eur J Radiol. 2016 Feb;85(2):366-72. doi: 10.1016/j.ejrad.2015.11.015. Epub 2015 Nov 15.

Abstract

Purpose: To assess the feasibility of virtual nonenhanced (VNE) images derived from single-source fast kVp-switching dual-energy CT, namely material suppressed iodine (MSI) images in evaluating gastric tumors, and to calculate potential radiation dose reductions for a dual-phase gemstone spectral imaging (GSI) abdominal CT compared with a standard triple-phase protocol.

Materials and methods: The institutional review board approved this prospective study; all patients provided written informed consent. 95 patients underwent preoperative enhanced CT that included true nonenhanced (TNE) with conventional 120 kVp mode, arterial phase (AP) and venous phase (VP) with GSI mode. MSI images were obtained from enhanced images of AP. Two abdominal radiologists independently evaluated the image quality of TNE and MSI images using a 5-point scale. Interobserver agreement was accessed by kappa value. The sensitivity for lesions on TNE and MSI images were evaluated. The CT number of various regions and image noise was measured on TNE and MSI images. The relative enhancement CT value of lesions, contrast to noise(CNR) of gastric tumors and effective radiation doses for triple-phase and dual-phase protocol were calculated. Differences were tested for statistical significance using Wilcoxon signed rank test and paired t-test.

Results: Interobserver agreement about image quality was excellent(all kappa>0.750). No significant difference in image quality was found between TNE 4.32 ± 0.53 and MSI 4.21±0.51 images (P>0.05). In respect to the sensitivities for gastric tumors, lymph node and liver metastasis, there were no significant differences between TNE and MSI images (all P>0.05). The CT number on TNE and MSI images, respectively, for gastric tumor were 32.79HU ± 6.45 and 35.02HU ± 5.01 (P=0.006); lymph nodes, 31.33HU ± 6.37 and 31.61HU ± 5.73 (P=0.334); liver metastasis, 37.79HU ± 8.40 and 39.76HU ± 8.50 (P=0.041); liver, 56.56HU ± 7.87 and 58.32HU ± 8.73 (P=0.057); psoas muscle, 52.30HU ± 5.63 and 51.80HU±6.86 (P=0.533); aorta, 38.16HU ± 4.91 and 42.49HU ± 4.49 (P<0.001). The CT relative enhancement value of gastric tumors, lymph node and liver metastasis had no significant difference between TNE and MSI images (P=0.083, 0.194, 0.156). MSI images had higher image noise than TNE images (P<0.001). The CNR of gastric tumor on MSI images was higher than that on TNE images (P<0.001). The effective dose for triple-phase was 25.1mSv±6.2 and that for dual-phase was 17.7mSv ± 4.1. The dose saving by removing the TNE was 7.5 mSv (30.5%).

Conclusion: In patients with gastric tumors, the MSI images derived from single-source fast kVp-switching dual-energy CT can provide comparable image quality to TNE images and reliable diagnostic information; replacing TNE images will lower radiation dose by 30.5%.

Keywords: Gastric tumor; Image quality; Radiation dose; Tomography; Virtual nonenhanced; X-ray computed.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Dose-Response Relationship, Radiation
  • Feasibility Studies
  • Female
  • Humans
  • Male
  • Middle Aged
  • Observer Variation
  • Prospective Studies
  • Radiography, Abdominal / methods*
  • Radiography, Dual-Energy Scanned Projection / methods*
  • Reproducibility of Results
  • Stomach Neoplasms / diagnostic imaging*
  • Tomography, X-Ray Computed / methods*