SU-D-217A-02: Effects of Enegy-Window Width and Spectral Effective Energy on Estimation of Gamma Camera Deadtime...

Assessment of count-rate performance (CRP) is part of routine quality assurance (QA) for gamma cameras. Manufacturers often specify deadtime (t) based on the NEMA decay method (NEMA 1-2007). The spectral conditions prescribed by NEMA are often difficult to duplicate during routine clinical QA testing and the appropriate energy window is poorly defined. The objective of this investigation is to evaluate the effects of energy-window selection and spectral conditions on estimates of t using the NEMA decay method.

CRP was evaluated intrinsically over a period of 48 hours with Tc-99m for two Siemens Symbia gamma camera detectors using the NEMA decay method. CRP measurement was repeated with varying amounts of scattering material (0-8 cm of acrylic) placed between the source and detectors to transform the incident spectrum. CRP measurements were also repeated using different photopeak-window widths (2-75%) as well as an open energy-window. The CRP data were fit to the paralyzable detector model via non-linear least-squares minimization to estimate t.

Estimates of t increased linearly with decreasing spectral effective energy. The effective energy was varied from 142-to-99 keV, which consequently altered the estimates of t by ∼0.095μs or 21% (p<0.01). Additionally, estimates of t increased as a power-law when the fraction of measured counts in the photopeak window decreased relative to open-window (total) counts. The ratio photopeak-window counts to open- window counts were varied from 1-to-0.12 which consequently altered the estimates of t by ∼8μs or 180%. The estimated change in t between 15% and 20% photopeak window is -0.13μs or 12.5%. Repeat measurements demonstrated that estimates of t using the decay method are precise to <2%.

Deadtime (t) determined using the decay method varied significantly with the spectral effective energy and photopeak-window width. Careful attention to measurement conditions are imperative when measuring t using the decay method for comparison against manufacturer's specifications or trending as part of routine QA program.