Program Information
Feasibility of Using Virtual Unenhanced Images to Replace True Unenhanced Images in Multiphase Renal CT Exams
D Popnoe*, S Zhou , C Ng , E Loyer , H Kang , H Kaur , A Jones , MD Anderson Cancer Center, Houston, TX
Presentations
MO-B-BRD-3 (Monday, March 9, 2015) 10:00 AM - 12:00 PM Room: Ballroom D
Purpose: To assess the feasibility of replacing true unenhanced (TUE) images with virtual unenhanced (VUE) images derived from dual-energy computed tomography (DECT) data in multi-phase renal CT exams, which could allow for reduced patient dose and increased patient throughput.
Methods: Virtual unenhanced scans were constructed from DECT data using a multi-material decomposition algorithm for 60 consecutively selected patient studies for a retrospective study. Image quality was assessed for VUE and TUE images qualitatively and the differences were tested with a mixed effects model. CT numbers were measured in the liver, spleen, spine, aorta, cystic lesions, subcutaneous fat, renal cortex and medulla and the differences between TUE and VUE measurements were tested with a Student’s paired t-test.
Results: Mean CT numbers (± standard deviation) on TUE and VUE images, respectively, were 51.16 ± 10.74 and 52.27 ± 10.16 (P < 0.001) for liver, 41.35 ± 6.35 and 47.62± 6.13 (P < 0.001) for spleen, 155.32 ± 61.91 and 82.15 ± 43.72 (P = 1.00) for spine, 35.91 ± 5.7 and 44.76 ± 9.44 (P = 0.543) for aorta, 15.97 ± 10.11 and 15.14 ± 10.93 (P < 0.001) for cystic lesions, -99.15 ± 15.96 and -95.64 ± 15.31 (P < 0.001) for subcutaneous fat, 27.45 ± 5.97 and 40.42 ± 7.41 (P = 0.999) for renal cortex, and 29.11 ± 4.79 and 28.1 ± 7.63 (P < 0.001) for renal medulla. VUE image quality was non-inferior to TUE images in all cases except in the visualization of major vessels and the depiction of the liver parenchyma.
Conclusion: This work indicates that although there is potential for VUE images to replace TUE images, there is still work to be done to identify the optimal set of image acquisition and processing parameters.
Funding Support, Disclosures, and Conflict of Interest: CABIR/GE In-kind research contribution study plan
Contact Email: