An investigation of accuracy of iterative reconstructions in quantitative SPECT

The iterative methods of image reconstructions represent the promising way for recovering the activity distribution from nuclear medicine acquisitions. One of the clinical areas where the accurate estimation of activity is extremely important is the radiotherapy of tumours. The goal of this study is to estimate the accuracy of the currently used in clinics iterative maximum likelihood expectation maximization (MLEM) methods in different realistic oncology situations. Numerical model was created to reproduce some of the clinical cases from the Internal Radiotherapy (IRT). Monte-Carlo simulations were utilized to generate sets of projections with the realistic noise level. The quantitative capability of the method was evaluated by performing a comparative analysis of true and reconstructed distributions. The influence of incorporation of physical effects (attenuation, scatter, and resolution loss) and algorithmic parameters (number of projections and iterations) on the solution accuracy and the convergence behaviour was studied.