Yu-Ling Huang MD, Presenter: Nothing to Disclose

Jui-Sheng Hsu MD, PhD, Abstract Co-Author: Nothing to Disclose

Ming-Lun Chiu MD, Abstract Co-Author: Nothing to Disclose

Ding-Kwo Wu MD, Abstract Co-Author: Nothing to Disclose

Chien-Hung Lee Lee, Abstract Co-Author: Nothing to Disclose

Wei-Chen Lin MD, Abstract Co-Author: Nothing to Disclose

Paul Ming-Chen Shih MD, Abstract Co-Author: Nothing to Disclose

Chiao-Yun Chen MD, Abstract Co-Author: Nothing to Disclose

To assess the diagnostic accuracy of HCC as well as the image quality, and radiation dose at low tube voltage with Sinogram Affirmed Iterative Reconstruction (SAFIRE) algorithm using dual-source, dual-energy CT (DECT)

We reviewed database of dual energy triple phase CT for liver tumor survey with DECT. Seventy patients (49 men, 21 women; age range, 40-84 years) with proven hepatocellular carcinoma (HCC) were enrolled. Source data of weighted average image (simulating 120kVp) and a pure 80 kVp data set were reconstructed with conventional filtered back projection (FBP) and SAFIRE. Totally, there were 8 protocols: 120 kVp with FBP (A, E), 120 kVp with SAFIRE (B, F), 80 kVp with FBP (C, G), and 80kVp with SAFIRE (D, H) at late arterial phase and portal venous phase, respectively. For these protocols, the following variables were evaluated: signal-to-noise ratio (SNR) of tumor, liver, paraspinal muscles, aorta, pancreas; contrast-to-noise ratio (CNR) of tumor; mean image noise; and effective dose. Two radiologists qualitatively scored the eight data sets regarding tumor detection and image quality. A repeated-measures ANOVA was performed for statistical analysis.

Compared with the conventional FBP data set at 120 kVp, the SAFIRE reconstruction data set collected at 80 kVp demonstrated significantly higher SNR (11.9 vs 9.5, protocol D vs A, P<0.001; 11.8 vs 9.4, protocol H vs E, P<0.001) and CNR (4.9 vs 3.0, protocol D vs A, P<0.001; 1.8 vs 1.6, protocol H vs E, P=0.0045) at the both vascular phases. In the arterial phase, the detection rate of the protocol D was better than that of protocol A (95% vs 90%). However, in the portal venous phase, no difference existed between the protocol H and E (85% vs 85%). In either 120 kVp or 80 kVp sets, compared with FBP, SAFIRE showed significantly lower image noise at the both vascular phases. The 80 kVp protocols significantly reduced the effective dose by 41%, as comparing with 120 kVp protocols.

Compared with weighted-average images plus FBP reconstruction, a low tube- voltage plus SAFIRE reconstruction can not only provide higher SNR, CNR, better lesion conspicuity during the arterial phase, but also reduces radiation dose.

For the patients with HCC, many follow-up imaging studies may be compulsory. Protocol using 80 kVp plus SAFIRE can reduce radiation dose in both vascular phases without compromise in tumor detection.

Huang, Y, Hsu, J, Chiu, M, Wu, D, Lee, C, Lin, W, Shih, P, Chen, C, Low Tube Voltage CT Using Sinogram Affirmed Iterative Reconstruction for Patients with Hepatocellular Carcinoma: Assessment of Diagnostic Accuracy, Image Quality, and Radiation Dose.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12028554.html