Effective analysis of image reconstruction algorithms in nuclear medicine using fuzzy PROMETHEE

Abstract

In medical imaging, the sets of projections obtained from either gamma radiation or X-radiation are converted into either a 2 dimensional (2D) or three dimensional (3D) images via a technique called image reconstruction. Nuclear medicine in particular employs this reconstruction technique to display the the image or radioisotope distribution within the patient's body. These techniques are integral part of the imaging modalities; they are complex computer algorithms which uses analytical or iterative methods to transform the generated projections into images. They come in various types and are characterized with certain features leading to the creation of image of diagnostic quality. Among these quality features are spatial resolution, variance, bias, contrast, noise etc. Availability of several reconstruction techniques makes it a huge task during the selection process of the best method for a particular modality. This study examines and compared the most commonly employed techniques using a Fuzzy-based Multi-criteria decision making method. This Fuzzy PROMETHEE method was employed in order to solve the decision making problems in nuclear imaging. Evaluation result showed that FBP is the alternative among the evaluated reconstruction techniques, due to its numerous distinct quality features. Furthermore, LM-OSEM was found to have the highest positive outranking flow.