The effect of energy and source location on gamma camera intrinsic and extrinsic spatial resolution: an experimental and Monte Carlo study

Quantification of nuclear medicine image data is a prerequisite for personalized absorbed dose calculations and quantitative biodistribution studies. The spatial response of a detector is a governing factor affecting the accuracy of image quantification, and the aim of this work was to model this impact. To simulate spatial response, a value for the intrinsic spatial resolution (Rintrinsic) of the gamma camera is needed. Rintrinsic for 99mTc was measured over the field of view (FOV) and an experimental setup was designed to measure Rintrinsic for radioisotopes with higher photon energies. Monte Carlo (MC) simulations, using the codes SIMIND and GATE, were used to investigate the extrinsic effect of Rintrinsic as a function of energy and its variation across the FOV. A method was developed to calculate energy-dependent blurring values for input to MC simulations, by separate consideration of the Compton scatter and photoelectric effect in the crystal and statistical variation in the signal. Inclusion of energy-specific blurring values in simulations showed excellent agreement with experimental measurements. The maximum pixel count rate can change by up to 18% when imaged at two different points in the FOV, and errors in the maximum pixel count rate of up to 11% were shown if a blurring value for 99mTc was used for simulations of 131I. We demonstrate that the accuracy of MC simulations of gamma cameras can be significantly improved by accounting for the effect of energy on intrinsic spatial resolution.