Tomoko Hyodo MD, Presenter: Nothing to Disclose

Yoshifumi Nakauchi MD,PhD, Abstract Co-Author: Nothing to Disclose

Koji Yamada RT, Abstract Co-Author: Nothing to Disclose

Masayuki Kudo PhD,RT, Abstract Co-Author: Employee, General Electric Company

Teruhito Mochizuki MD, Abstract Co-Author: Nothing to Disclose

Takamichi Murakami MD, PhD, Abstract Co-Author: Nothing to Disclose

Masahiro Okada MD, Abstract Co-Author: Nothing to Disclose

Kazunari Ishii MD, Abstract Co-Author: Nothing to Disclose

To assess the contribution of high definition scan mode, high-resolution convolution kernel and adaptive statistical iterative reconstruction (ASIR) technique to the luminal visualization of coronary artery stents using 64-MDCT.

Two different stent designs each with two diameters (2.5 and 3.5 mm) in simulated vessels were filled with iodinated contrast, fixed around a container submerged in iodinated contrast (simulated the left ventricle), and scanned. The sampling rates were set to 656 views for standard scan mode (S-St) and 1662 views for high definition scan mode (S-HD). Data sets were reconstructed with both a standard (K-St) and a high-resolution convolution kernel, while the latter applied conventional filtered back projection (FBP) alone (K-HR) or two levels of ASIR-FBP blending (50:50 [K-HR/A50] and 100:0 [K-HR/A100]). Each data set was evaluated for 1) the in-stent luminal diameter (ISLD) based on the full width half maximum of the attenuation profile across the stent, 2) the maximal luminal cross-sectional area (CSA) obtained by reformatting the longitudinal view of each stent, and 3) the mean CT value and its standard deviation (SD: image noise) of a region of interest on the container. The data was analyzed by repeated-measures ANOVA with post hoc Scheffe's test.

For each scan mode, the ISLDs in the three conditions of K-HR, K-HR/A50 and K-HR/A100 were similar and larger than in K-St. The CSAs resulted in the same tendency of the ISLDs. ANOVA showed a significant effect of the reconstruction kernel on both ISLD and CSA, while the effect of the scan mode was not significant. The interaction between the scan mod and the reconstruction kernel was significant for ISLD. Post hoc analysis showed that both ISLD and CSA improved by integrated use of the high definition scan mode and high-resolution convolution kernel compared to use of each of the two (not significant). The mean CT values within the container didn’t vary with the reconstruction kernel, whereas adding ASIR decreased SD.

Integrated use of the high definition scan and high-resolution convolution kernel contributes to intra-stent visualization. ASIR technique lowers noise maintaining high resolution.

ASIR with high definition scan and/or high-resolution convolution kernel improves visualization of intraluminal coronary artery stent by lowering noise without any degradation of MDCT image.

Hyodo, T, Nakauchi, Y, Yamada, K, Kudo, M, Mochizuki, T, Murakami, T, Okada, M, Ishii, K, In Vitro Evaluation of Coronary Artery Stents by 64ch MDCT: Improvement in Luminal Visualization Using High Definition Scan Mode with Adaptive Statistical Iterative Reconstruction Technique.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11034199.html