Design optimization and simulated performance evaluation of a self-collimated cardiac SPECT with simultaneous high sensitivity and resolution.

Abstract

Background: Cardiac single photon emission computed tomography (SPECT) is an important noninvasive molecular imaging technology for the diagnosis and risk stratification of heart diseases. However, since its inception, the clinical impact of cardiac SPECT has been constrained by its reliance on mechanical collimation using a metal apertured-plate structure, which causes an inherent inverse interdependency between the system's resolution and sensitivity. Recently, our group introduced self-collimation (SC) detector architecture in single photon emission imaging: in a structure with multiple layers of sparsely distributed active detector elements, certain detectors also function as collimators for others. We have shown that SPECT with SC detector architecture transcends the performance of its counterpart with conventional mechanical collimation.

Purpose: This study aims to (1) produce the blueprint of a cardiac SC-SPECT system that achieves high sensitivity and resolution simultaneously, and (2) evaluate the system's performance through Monte Carlo simulation studies.

Methods: Based on our prior SC-SPECT studies and general design considerations for cardiac imaging, we choose a half-hexagon configuration, comprising three identical trapezoid-shaped detector heads, for the detector gantry of the cardiac SC-SPECT. The targeted field of view (FOV) is a spherical volume with a diameter of 190 mm. Each detector head includes a tungsten plate on the side facing the FOV, followed by four stacked detector layers with spacing between each layer. With the distance from the fourth detector layer to the center of FOV, as well as the size of individual detector scintillators fixed, we use an established system resolution and variance evaluation scheme to optimize detector configuration parameters, namely, the placement positions of the first three detector layers, the distribution pattern of the apertures and aperture to metal ratio on the tungsten plate. The SC-SPECT with the set of optimized configuration parameters is then evaluated with a set of simulated phantom studies, and the results are compared against that of a conventional dual-head SPECT system. The phantoms include a hot-rod phantom, a cold-rod phantom, and a XCAT phantom configured for emulating various myocardial ischemia conditions under realistic injection dose and variable acquisition times.

Results: A cardiac SC-SPECT design with optimized parameters is obtained. The system achieves an average sensitivity of 0.54% within the FOV, and can clearly resolve hot rods with a diameter of 4 mm and cold rods with a diameter of 5 mm. In contrast, the conventional SPECT system exhibits an average sensitivity of 0.02% in the FOV, resolves 6 mm diameter hot rods and none of the cold rods (4-9 mm) with a body-contour orbit, and resolves 5 mm diameter hot rods and 7 mm diameter cold rods with a 150 mm radius circular orbit. The cardiac SC-SPECT reliably identifies cardiac defects with significantly shorter time than that required by its conventional SPECT reference.

Conclusion: The optimized cardiac SC-SPECT shows very promising performance improvement over conventional SPECT.