A simulation study of a long axial field of view whole-body PET scanner using cylindrical and anthropomorphic phantoms

Clinical whole-body PET scanners mostly have an axial field of view (AFOV) ranging from 15–21 cm in length, which limits sensitivity and prevents the acquisition of whole-body dynamic images. We present results from a simulation study for a fully 3D, 2 meter long AFOV whole body PET scanner where we use cylindrical and anthropomorphic phantoms to characterize the count-rate performance using noise equivalent counts (NEC) as the metric. For whole-body scanners with a conventional AFOV, the NEC is generally expected to increase when data from larger ring differences is accepted. However, the lines of response with large ring difference values may encounter a large amount of tissue and attenuation, which may limit the number of trues detected without a concomitant reduction in scatter and randoms, thus limiting NEC. When the activity level in the phantom increases, the dead-time and randoms also increase significantly, eventually reducing the NEC performance as well. In this simulation study, the NEC was computed as a function of maximum acquired ring difference and activity for each phantom to find optimal performance parameters for our proposed scanner. Our simulation results show our proposed scanner performs 25–50 times better than existing clinical scanners in terms of NEC performance. With significant gains in sensitivity, a long AFOV scanner can potentially be used for whole body dynamic PET studies.