Optimization of an iterative SPECT reconstruction algorithm utilizing a partial volume effect correction method

Single Photon Emission Computed Tomography (SPECT) is a widely used imaging technique in nuclear medicine. In SPECT imaging the Maximum Likelihood Expectation Maximization (MLEM) algorithm is a popular reconstruction technique, although the MLEM based image reconstruction is a time consuming process especially in the case if we use it with attenuation correction and compensation for the distance dependent spatial resolution of the detector. In this research we have developed a new MLEM based iterative SPECT reconstruction algorithm reducing significantly the reconstruction time. The new reconstruction method is divided into two phases. In the phases different spatial resolution of the reconstructed volume is used. In the first phase of the reconstruction we scaled down the original spatial resolution of the detected images in order to decrease the amount of data to process and - simultaneously - decrease the computational complexity of the reconstruction algorithm. In the second phase we increased the spatial resolution of the reconstructed image to the original one. In this phase we applied reverse diffusion based partial volume effect (PVE) correction to partially compensate the negative effects of the low resolution. The method has been tested with mathematical phantoms (with a bullet phantom and with the NCAT phantom) with additional noise. Convergence curves shows that the suggested method is able to reconstruct with similar quality than the high resolution reconstruction while the execution time of the algorithm has been significantly reduced.