Masters Theses

Ming Xu

Date of Award

5-1992

Degree Type

Thesis

Degree Name

Master of Science

Major

Electrical Engineering

Major Professor

Dagana Brzakovic

Committee Members

George Kabalka

Abstract

The purpose of this research was to develop an effective method for attenuation correction for Positron Emission Tomography (PET). This involved the study of the principle of the attenuation problem, methodology of attenuation correction by using transmission scan images, image processing and segmentation of transmission images and software development for the graphic user interface used to fulfill the procedure. Attenuation is the predominant quality problem of quantitative PET studies. Conventional attenuation correction methods use either predefined geometrical attenuation material shapes (calculated method), or directly use the transmission scan image for attenuation correction (measured method). Due to the simplicity of these methods, poor results are inevitable. Attenuation problem thus makes clinical PET diagnosis even more difficult. The Adaptive Attenuation Correction Tool introduced here is a complete, user-friendly graphic user interface package. This tool provides an improved measured attenuation correction method by combining transmission image processing and interactive boundary redrawing options. The algorithm of image processing on the transmission scan was developed by modeling the low-count transmission and noise distribution. In addition to the image processing of the transmission scan images, the tool also provides interactive options and monitors intermediate results which gives the operators more control of attenuation correction throughout the whole process. A comparison of this tool with the conventional methods reveals significant improvements. It requires much shorter transmission scanning time while producing better attenuation correction results. Clinical use of the technology should result in a dramatic reduction of patient bed-time and potentially better quantitative PET studies.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS