ISSN: 2155-9880
Alexandru Naum, Nina Kleven-Madsen, Martin Biermann, Boel Johnsen, Birger A Tvedt, Svein Rotevatn, Jan Erik Nordrehaug and Tore Bach-Gansmo
Objective: To investigate the impact of a reconstruction method incorporating corrections for scatter, attenuation, and distance dependent detector response on cardiac single emission computed tomography (SPECT) perfusion studies compared with filtered back projection (FBP).
Materials and methods: A total of 20 patients underwent same-day, rest-stress SPECT/CT myocardial perfusion imaging. Images were reconstructed using iterative 3D ordered-subsets expectation-maximization (OSEM 3D) and FBP algorithms. Stress and rest myocardial perfusion defects were quantified using polar maps and normal database comparisons. The Bland-Altman plots were used to assess their degree of agreement. Results were confirmed by coronary angiography. The contrast, contrast to noise ratio and signal to noise ratio were used for the quantitative evaluation of the reconstruction quality.
Results: Perfusion defect extent quantification on OSEM 3D reconstructed images agreed and correlated well with defect extent quantification on FBP reconstructed images (bias ± standard deviation, -15% ± 20; r = 0.63) during stress and at rest (-10% ±15; r=0.70). Agreement and correlation were similar for severity scores during stress (-1.02 ± 1.77 SD’s; r=0.62), and at rest (-1.10 ± 1.49 SD’s; r=0.61). There were no statistically significant differences between the methods regarding perfusion defect extent or severity. The overall agreement rate with coronary angiography was similar. OSEM 3D reconstruction algorithm significantly increases the image contrast by 31% (P<0.05).
Conclusions: Compensation for detector response, attenuation and scatter improves image contrast compared with FBP. Applying quantitative analysis, OSEM 3D reconstruction produced increased image contrast compared to FBP, but similar results regarding the size and severity of left ventricular perfusion defects.