Iterative methods of tomographic image reconstruction in medical imaging

摘要

An investigation is conducted by four iterative image reconstruction methods, including the algebraic reconstruction technique (ART), the iterative least-squares technique (ILST), and Richardson's iterative method of solving a system of equations. An alternative to ILST, called the iterative averaging technique (IAT), is presented. Each method is written in an iterative matrix form. Each can be used to reconstruct the image in a single pass and is optimum for some criterion. Weighting matrix properties and their effects on the reconstruction matrix, number of reconstruction operations, and computation time requirements are noted.