Evaluation of collimator choice and scatter correction on 123I SPECT images

Abstract

Aim To compare the performance of low-energy (LE) and medium-energy (ME) collimators and the use of scatter correction on 123I SPECT images as assessed by visual inspection and semi-quantitative analysis. Materials and methods Both brain and heart/thorax phantoms were used to evaluate the effects of collimator selection and correction methods on brain and cardiac 123I imaging. Experiments were performed to measure the target-to-background ratio for calculation of object contrast. The projection data were corrected using X-ray-based attenuation maps for attenuation correction (AC) and dual-energy window method for scatter correction (SC). Reconstructed images using different methods were compared, which included FBP (filtered backprojection), and OSEM (ordered subsets expectation maximization) without any correction, with AC, with SC, and with AC and SC. Results In both phantom studies, the image contrasts provided by the ME collimator were better than that of the LE collimator. The OSEM with simple dual-energy window SC method improved image contrast and visual image quality as compared to the FBP method. The use of AC improved the visual image quality but did not improve image contrast, if SC method was not implemented at the same time. Conclusions This study demonstrates the importance of SC and ME collimator in 123I SPECT imaging. If image contrast is the figure of merit, then SC should be performed and the use of the ME collimator appears to be preferable in semi-quantitative 123I SEPCT studies.