Computer evaluation of a novel multipinhole SPECT system for cardiac imaging

Multipinhole collimation for clinical myocardial perfusion imaging dates back to the 1970’s and 1980’s. However, the acceptance and use of pinhole collimator technique for application in single photon emission computed tomography (SPECT) was impeded by wide-spread availability of use for rotational dual-head cameras for general purpose imaging. Today virtually all SPECT imaging is performed with a parallel-hole collimator. In the past few years, experimental small animal SPECT systems have seen progress with pinhole collimation. There has also been an interest in designing dedicated cardiac SPECT systems for improved sensitivity to reduce imaging time. This has renewed clinical interest in multipinhole radionuclide cardiac imaging. It is known that pinhole collimation surpasses parallel-hole collimation in both sensitivity and resolution for small objects located close to the pinhole aperture. However, the short object-to-detector distance limits the size of the field of view (FOV) for the pinhole collimator, which causes part of the emission rays to miss the detector. The lost data are said to be truncated from the projection measurements. The objective of this work is to investigate the effect of truncation and limited angular sampling on image quality for a novel cardiac imaging system with stationary multipinhole collimation. Computer simulations show that the system produces high quality images of the myocardium comparable to that of a rotating multipinhole detector system. Improved quantitation for truncation of the interior problem is accomplished by calibrating the system matrix using a uniform phantom in the FOV of known activity.