Hot sphere detection limits for a dual head coincidence imaging system

Phantom studies were performed to investigate detectability limits for a dual head coincidence imaging system. Spheres of 2.8, 2.2, 1.6, 1.2, 1.0, 0.8 cm inner diameter were placed at a radius of 5.75 cm in a 22 cm diameter cylindrical phantom. The contrast to noise ratio (CNR) of the spheres to background was used as the authors' metric of detectability. CNR values were determined for three different sphere to background ratios (10:1, 5:1, and 3:1); for three different background count densities (high, medium, and low); and for three different energy window requirements (photopeak only, photopeak and photopeak-Compton, and all combinations). Based upon the authors' visual observation, the smallest detectable spheres were 0.8 cm, 1.2 cm and 1.6 cm for a 2.5 million event photopeak-Compton image set with sphere to background ratios of 10:1, 5:1 and 3:1, respectively. The CNR values for these spheres were 5.1, 7.8, and 4.8. Over all observations, spheres with a CNR greater than 4 were detectable.