The Added Value of Virtual Unenhanced Images Obtained From Dual-energy CT Urography in the Detection and Measurement of Urinary Stone

Abstract

To evaluate the detection and quantification of urinary stones using virtual unenhanced images (VUE) at different phases and slice thickness in contrast-enhanced dual-energy CT Urography (DECTU) in comparison with true unenhanced images (TUE). One hundred and twelve urinary stone patients who required triphasic DECTU were analyzed. Data were reconstructed as the followings: TUE images with 1.25 mm thickness (TUE portal venous phase VUE images with 1.25 and 5-mm thickness (VUE(VP)_1.25 mm and VUE(VP)_5 mm) and excretory phase VUE images with 1.25 and 5-mm thickness (VUE(EP)_1.25mm and VUE(EP)_5mm). The. The stones were divided into large (≥5 mm) and small stones. The detection rate, size and CT value of stones were assessed by 2 radiologists and statistically compared among the above groups. Two hundred and thirty urinary stones (163 large and 67 small stones) were detected on TUE_1.25 mm images. For large stones, the detection rate on VUE(VP)_1.25 mm, VUE(VP)_5 mm, VUE(EP)_1.25 mm and VUE(EP)_5 mm was 100%, 96.9%, 85.9%, and 80.4%; while for small stones, the rate was 77.6%, 37.3%, 46.3%, and 23.9%, respectively. VUE(VP) images significantly improved the stone detection rate compared with VUE(EP) images at both slice thicknesses. In general, VUE images identified stones with smaller sizes and lower Hounsfield units, but thinner slice thickness images reduced the inaccuracy. Inter-reader agreement of the stone detection revealed a k value range from 0.85 to 0.94 for TUE and VUE images. Large stones (≥5 mm) can reliably be detected on thin section VUE(VP) images with 33% radiation dose reduction. However, for small stones TUE remains superior. Stone size is underestimated on VUE images.