Single source split filter dual energy: Image quality and liver lesion detection in abdominal CT. Issue 126 (May 2020)

• Record Type: Journal Article
• Authors: Eichler, Mathis; May, Matthias; Wiesmueller, Marco; Saake, Marc; Heiss, Rafael; Uder, Michael; Wuest, Wolfgang
• Is Part Of: European journal of radiology. Issue 126(2020)
• Publication Date: 2020-05
• Subjects: A Mean attenuation -- CNR Contrast to noise ratio -- DI Default image -- DLP Dose length product -- ED Effective radiation dose -- HU Hounsfield units -- keV Kilo electron volt -- N Image noise -- ROIs Region of interest -- VMI Virtual monoenergetic images; Abdominal imaging -- Breast carcinoma -- Single source dual energy; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2020.108913
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: 40 keV delivers highest CNR for vessels and organ parenchyma. Newly induced artifacts are exclusively found at low keV reconstructions. Overall image quality and assessment of liver lesions is best at 80 keV. Abstract: Purpose: Aim of this study was to evaluate image quality of single source dual energy CT (SSDE-CT) using split filter technique in oncologic abdominal CT. Method: 51 consecutive patients with newly diagnosed breast carcinoma were prospectively enrolled in this study and underwent a staging examination of the abdomen using SSDE-CT (120 kV, split filter technique, 400 ref. mAs). Inline default images (DI) and post-processed virtual monoenergetic images at 40 keV, 50 keV, 60 keV, 70 keV and 80 keV were reconstructed. Objective image quality was evaluated as contrast to noise ratio (CNR) for liver parenchyma, portal vein, spleen, pancreas, aorta and hypoattenuating liver lesions. Subjective image quality was rated on a 5-point scale. Image quality at different keV settings was analyzed in paired t -tests. Results: CNR was highest at 40 keV for vessels (portal vein: 9.0, aorta: 8.8, all p < 0.001) and for upper abdominal organs (spleen: 4.8, all p < 0.001; pancreas: 2.7, all p < 0.01 except p = 0.93 for 50 keV; liver parenchyma: 3.4, all p < 0.01). Highest CNR values for hypoattenuating liver lesions were found at 40 keV (7.7, all p < 0.001). Overall subjective image quality was highest with 80 keV and DI (both 4.8, all p < 0.001). Artifacts were most pronounced at 40 keV. Conclusions: High image quality can be obtained with SSDE-CT of the abdomen. Lowest monoenergetic reconstructions provide the highest image contrast and should be used for vessel evaluation. The best trade-off between artifacts and parenchymal contrast can be obtained with 80 keV images.