Statistical Tomographic Reconstruction in Nuclear Medicine

Abstract

In nuclear medicine, an estimate of the three-dimensional spatial distribution of injected radionuclide is needed for diagnosis. This estimate is obtained through a reconstruction algorithm. The tomographic reconstruction in nuclear medicine can be divided into two groups: analytical and statistical methods. It has been shown that statistical image reconstruction outperforms analytical reconstruction methods since the data are noisy and follow Poisson distribution. In addition, statistical Bayesian image reconstructions are capable of incorporating accurate system models and including a penalty (prior) to capture the known characteristics of the underlying object. We will introduce the imaging model, and discuss the problems of statistical image reconstruction in nuclear medicine.