Performance evaluation of resolution and sensitivity of C‐SPECT's variable slat‐stack collimator

Background Myocardial perfusion imaging is commonly performed using SPECT, where both general‐purpose and dedicated scanners are available. A limitation with general‐purpose systems has been the inability to image dynamically since different projections are obtained far apart in time due to scanner rotation. Dedicated systems can have this capability since they acquire completely sampled projections (i.e., those with enough angular views for reconstruction) with short time frames. C‐SPECT, does not need any scanner or patient motion to obtain complete projections, allowing fast dynamics. When imaging fast dynamics, the optimal collimator settings are not necessarily the same as for static imaging, where longer acquisitions can be utilized. Thus, C‐SPECT offers adaptive collimation in the transverse and axial directions. Purpose The performance of adaptation in the axial direction was characterized herein. Methods The ratio of the resolution metric in high‐sensitivity mode to that in the high‐resolution mode, termed resolution boost factor, was determined. Analogously, the sensitivity boost factor was also determined. Comparisons were made with theory and simulations. Results The boost factors for resolution and sensitivity, averaged over the 14 modules of the system, were determined to be 1.72 and 1.75, respectively. Conclusions The boost factors, which ideally would be two, were between 10% and 15% below optimal values and tracked each other, suggesting mechanical challenges in the apparatus, such as incomplete closure of adjacent slats, but show reasonably successful adaptation between modes.