Indirect computed tomography venography of the lower extremities using single-source dual-energy computed tomography: Advantage of low-kiloelectron volt monochromatic images

Purpose: To study the performance of dual-energy indirect computed tomography (CT) venography from single-source dual-energy CT in the assessment of lower extremity deep venous thrombosis (DVT). Materials and Methods: In a retrospective study, 110 patients suspected to have pulmonary embolism (PE) underwent dual-energy CT venography on a single-source dual-energy CT scanner as a part of CT pulmonary angiography protocol at 3 minutes after injection of contrast material. Two radiologists evaluated 50-kiloelectron volt (keV) and 70-keV monochromatic images reconstructed from a dual-energy CT scan for image quality, image noise, venous contrast, and confidence level in interpretation for DVT using a scale of 1-5. In addition, a combined 50-keV and 70-keV data set was assessed for confidence level in image interpretation. Attenuation, contrast-to-noise ratio (CNR), and objective noise were measured in bilateral common femoral and popliteal veins. Data were analyzed using Student t test and Wilcoxon rank sum test. Radiation dose was measured for dual-energy CT venography protocol. Results: A diagnosis of DVT was made in 8 of 110 patients (7.27%). The subjective image quality was comparable between 50-keV and 70-keV images (4.3 vs 4.5; P > .05). The subjective venous contrast opacification (4.7 vs 3.5; P = .0036) and confidence (4.8 vs 3.9; P = .0028) in image interpretation were superior at 50 keV. Confidence level for interpretation on combined 50-keV and 70-keV series (score 4.7) was similar to that for 50-keV series (score 4.8). Compared with 70-keV data, 50-keV data yielded 90% increase in intravascular CT attenuation (207.4 Hounsfield units [HU] ± 39.0 vs 106.8 HU ±7.6; P <.0001) and higher CNR (10.7 ± 4.07 vs 7.2 ± 4.1; P = .0001) of the deep veins. However, objective noise at 50 keV was higher (14.8 HU vs 6.5 HU; P = .0031). Because of inadequate contrast opacification, 6% of CT venography studies were deemed suboptimal for rendering a diagnostic interpretation on 70-keV images, but these images were considered acceptable at 50 keV. The mean effective radiation dose for the dual-energy CT venography examination was 4.2 mSv. Conclusions: Optimal image quality with substantially higher venous attenuation is provided by 50-keV monochromatic images from dual-energy CT venography acquisition compared with 70-keV images. The 50-keV monochromatic images increase the confidence in the image interpretation of DVT and decrease the number of indeterminate studies.