Design study of a cardiac-dedicated PET system

Abstract

The work studies the feasibility of developing a cardiac-dedicated PET system using Monte Carlo simulation tools. The proposed system comprises a dual-panel geometry and both panels utilize fine crystal elements (~2–3mm) to help improve high spatial resolution. The system performances were studied with respect to photon detection sensitivity, count-rate performances, spatial resolution, depth-of-interaction (DOI) capability, and potential contrast recovery improvement due to time-of-flight (TOF). For a 2:3 dual panel configuration (front panel: 24.0×18.0cm2 and back panel: 36.0×27.0cm2), the system is able to achieve a peak photon sensitivity of ~13.0% at the location of the heart. For count-rate performances, the system is able to yield a continuous increase of noise equivalent count rate (NECR) as a function of the total activity of radiotracers between 10mCi and 50mCi. When crystal elements of 2×2×20cm3 are deployed in the front panel and crystal elements of 3×3×20cm3 are deployed in the back panel, the proposed system is not only able to revolve 2mm diameter spheres in the image slices parallel to the panels but also demonstrates good resolution uniformity of less than 5%. On the other hand, significant resolution degradation occurs in the direction perpendicular to the panels due to the following two factors: limited angular coverage and finite DOI resolution. Such degradation was quantitatively analyzed for five different DOI resolutions (0mm, 2.5mm, 5mm, 10mm and 20mm), as well as three different crystal configurations. Finally, the contrast study with a heart-like phantom (comprising lumen, aortic wall, myocardium and small lesions) indicates that TOF capability (time resolution: 250–500ps) does not significantly improve image quality and lesion detectability, in terms of contrast-to-noise ratio (CNR).