GATE Simulations of Small Animal SPECT for Determination of Scatter Fraction as a Function of Object Size

In human emission tomography, combined PET/CT and SPECT/CT cameras provide accurate attenuation maps for sophisticated scatter and attenuation corrections. Having proven their potential, these scanners are being adapted for small animal imaging using similar correction approaches. However, attenuation and scatter effects in small animal imaging are substantially less than in human imaging. Hence, the value of sophisticated corrections is not obvious for small animal imaging considering the additional cost and complexity of these methods. In this study, using GATE Monte Carlo package, we simulated the Inveon small animal SPECT (single pinhole collimator) scanner to find the scatter fractions of various sizes of the NEMA-mouse (diameter: 2-5.5 cm , length: 7 cm), NEMA-rat (diameter: 3-5.5 cm, length: 15 cm) and MOBY (diameter: 2.1-5.5 cm, length: 3.5-9.1 cm) phantoms. The simulations were performed for three radionuclides commonly used in small animal SPECT studies:
^99m
Tc (140 keV),
¹¹¹
In (171 keV 90% and 245 keV 94%) and
¹²⁵
I (effective 27.5 keV). For the MOBY phantoms, the total Compton scatter fractions ranged (over the range of phantom sizes) from 4-10% for
^99m
Tc (126-154 keV), 7-16% for
¹¹¹
In (154-188 keV), 3-7% for
¹¹¹
In (220-270 keV) and 17-30% for
¹²⁵
I (15-45 keV) including the scatter contributions from the tungsten collimator, lead shield and air (inside and outside the camera heads). For the NEMA-rat phantoms, the scatter fractions ranged from 10-15% (
^99m
Tc), 17-23%
¹¹¹
In: 154-188 keV), 8-12% (
¹¹¹
In: 220-270 keV) and 32-40% (
¹²⁵
I). Our results suggest that energy window methods based on solely emission data are sufficient for all mouse and most rat studies for
^99m
Tc and
¹¹¹
In. However, more sophisticated methods may be needed for
¹²⁵
I.