Model-based Iterative Reconstruction for Multi–Detector Row CT Assessment of the Adamkiewicz Artery

Abstract

To determine if model-based iterative reconstruction (MBIR) can improve visualization of the Adamkiewicz artery on multi-detector row computed tomographic (CT) images compared with adaptive statistical iterative reconstruction (ASIR) and filtered back projection (FBP). This retrospective study was approved by the institutional review board, and written informed consent for the CT examination was obtained. Thirty-three patients underwent contrast material-enhanced 64-section multi-detector row CT for assessment of aortic aneurysm or dissection. Helical data were reconstructed by using FBP, ASIR, and MBIR. The signal-to-noise ratio of the aorta and contrast-to-noise ratio of the anterior spinal artery relative to the spinal cord were measured on multiplanar reformatted images. Visualization of the Adamkiewicz artery and its continuity with the intercostal or lumbar artery were evaluated by using a four-point scale. All image analyses were performed by two blinded, independent observers. The one-way analysis of variance and the Wilcoxon signed-rank test were used for statistical analysis. MBIR showed significantly better signal-to-noise and contrast-to-noise ratios than did ASIR and FBP (P < .05 for all comparisons) with good interobserver agreement (intraclass correlation coefficient of 0.93 for signal-to-noise ratio and 0.75 for contrast-to-noise ratio). The visualization score of the Adamkiewicz artery was also significantly better when MBIR was used (3.4 ± 0.8 and 3.6 ± 0.7 for observers A and B, respectively) than when ASIR (2.7 ± 1.1 and 3.0 ± 1.0, respectively) or FBP (2.5 ± 1.2 and 3.1 ± 0.9, respectively) was used. Use of the MBIR algorithm led to improved multi-detector row CT visualization of the Adamkiewicz artery when compared with the use of ASIR and FBP.