TY - JOUR T1 - Performance as a function of axial length for the scalable PennPET Explorer. JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 144 LP - 144 VL - 61 IS - supplement 1 AU - Margaret Daube-Witherspoon AU - Varsha Viswanath AU - Matthew Werner AU - Suleman Surti AU - Joel Karp Y1 - 2020/05/01 UR - http://jnm.snmjournals.org/content/61/supplement_1/144.abstract N2 - 144Introduction: The University of Pennsylvania has built the PennPET Explorer, a 70-cm axial length scanner, and demonstrated the benefits of long axial field-of-view (AFOV) systems. We are currently extending the scanner from 3 rings to 6 rings, doubling the axial length to 140 cm. Extending the length of the scanner will permit more of the subject to be imaged simultaneously with high sensitivity; however, the performance gains need to be understood and justified relative to the increased cost. The computational challenges and expected performance gains associated with doubling the AFOV of the PennPET Explorer have been addressed using simulated data. The practical challenges are mitigated by the scalable design and parallel architecture of data acquisition (DAQ). The purpose of this work is to characterize through simulation and measurement the gains in performance and possible trade-offs for human and large animal imaging as the axial length of the PennPET Explorer is extended. Methods: The PennPET Explorer is scalable; each ring of LYSO-based digital detectors has an axial length of 22.9 cm with 1-cm gaps between rings, and the rings are stacked to form scanners of various axial lengths. GATE simulations were performed of a scanner with 3, 4, and 6 detector rings to compare performance. Measurements are underway on the PennPET Explorer as rings of detectors are added to create the 6-ring system. Results: The axial resolution is relatively uniform (within 0.7 mm FWHM) throughout the AFOV for 3- and 6-ring systems even with an unrestricted axial acceptance angle. The total sensitivity for a 70-cm line source in air is predicted to increase from 97 kcps/MBq (3 rings) to 315 kcps/MBq (6 rings). The peak sensitivity in air at the center of the AFOV increases by 38% for 6 rings, but this increase is reduced with attenuation of oblique events by the body. With attenuation, the axial sensitivity profile is flat for the 6 rings over an extended axial range (~100 cm), implying uniform noise performance across much of the AFOV; this is important for single-bed imaging. The trues count rate is linear with activity as rings are added due to the parallel DAQ architecture; the noise equivalent count rate continues to increase beyond activities that will be encountered for human imaging, even first-pass cardiac studies. Contrast recovery does not vary across the AFOV, consistent with the very small losses in spatial resolution, due to attenuation of oblique events as the axial angle increases. Finally, the measured gains in lesion detectability near the center of the AFOV for a 3-ring system are expected to apply to a longer axial range as the AFOV increases and more of the body is imaged at high sensitivity. Conclusions: Measured and simulated results from the PennPET Explorer with a range of axial lengths from 70 to 140 cm demonstrate the overall performance benefits of the longer AFOV with very small or insignificant losses in axial resolution, although there are challenges with applying the NEMA measurements to systems with AFOVs beyond 70 cm. The clinical benefits of the 140-cm PennPET Explorer will be determined once this and other long AFOV scanners have been in use for longer periods of time. ER -