Validation of Convolution Based Forced Detection SIMIND with Analytical Collimator Response Modeling Using Gate

Monte Carlo is an important and well established research tool in emission tomography. While used extensively in research applications, Monte Carlo techniques are not typically implemented clinically due to is its low detection efficiency and resultant long acquisition times. In order to make this notoriously slow computational tool faster, variance reduction techniques known as convolution based forced detection (CFD) have previously been implemented into SIMIND MC code (CFD-SIMIND) by our group. In this method, at each interaction site within the object, photons are forced to travel in a direction perpendicular to the detector and are convolved with a Gaussian blurring kernel based on the collimator response. This study focuses on the feasibility of using CFD-SIMIND in the absence of septal penetration and collimator scatter modeling for low and medium energy isotopes. In this study, the results of CFD-SIMIND have been compared with GATE Monte Carlo in order to validate the CFD-SIMIND results. For this purpose a number of point source simulations were performed using CFD-SIMIND and GATE with low and medium energy isotopes (i.e. 99rnTc and 111 In respectively). 99mTc point source data (with low and medium energy collimator) and 111 In (with medium energy collimator) at different source to collimator distances in air and water was acquired. The point source data was then used for the validation of CFD-SIMIND in terms of spatial resolution, sensitivity, image profiles and energy spectra. In order to validate the results for extended source distribution and non-uniform attenuation media, XCAT phantom simulations are being performed. The 99m Tc data with low energy collimator, available so far, has also been presented in the study. A comparison similar to point source data obtained in water was also performed in this case. It has been observed that CFD-SIMIND correlates well with GATE for simple point source simulation data in the case of 99m Tc and 111 In even in the absence of collimator scatter and septal penetration modeling. Point source as well as XCAT phantom data for 99rnTc with low energy collimator showed that the use of CFD-SIMIND is valid for the isotope/collimator combination. In order to generalize the results, further simulation studies for other isotope/collimator pairs, used in point source simulations, are being performed for XCAT phantom.