SPECT collimator dependency of scatter and validation of transmission-dependent scatter compensation methodologies

Scatter correction is important for absolute quantification in SPECT. Among the approaches taken to achieve scatter correction, some methods using transmission information for scatter estimation require the scatter models based on the physical characteristics of the scattered photons. The influence of the SPECT collimator on the characteristics of the scattered photon, however, has not been well studied yet. In this paper, we investigate the relationship between scatter and collimator design for the use of the transmission-dependent convolution subtraction (TDCS) method. Monte Carlo simulations with various collimator acceptance angles and experiments with four collimators of two SPECT cameras were performed to obtain the scatter fraction (SF) and the line-spread function (LSF) of Tc-99m. The experiments with I-123 also used six collimators of three SPECT systems to obtain the SF and the LSF. In the simulations, the SF of Tc-99m did not change with the collimator acceptance angles. The LSFs measured for Tc-99m showed an excellent agreement between collimators. In the experiments for I-123, the SF was practically the same for the six collimators, although a small difference could be observed due to the septal penetration. Also, the shape of the LSF was very similar between the collimators, and the counts in the background region of reconstructed images did not differ when using different sets of TDCS parameters. These results reveal that scatter is independent of SPECT collimator design for Tc-99m even I-123, and that the TDCS could be applicable to scatter correction of I-123 using a unique collimator-independent set of parameters.