Identifier

etd-04132005-163806

Degree

Master of Science (MS)

Department

Physics and Astronomy

Document Type

Thesis

Abstract

Radiotherapy treatment planning utilizing PET and CT is rapidly gaining acceptance in oncology. A limiting factor of the dual modality is the PET/CT alignment. A small error in PET/CT alignment may result in giving large doses of radiation to healthy tissues as a result of poor treatment planning. For this purpose, regular quality assurance testing of PET/CT must be performed. Separate QA procedures and phantoms have been developed for the two different modalities. In particular, many existing phantoms cannot be used for both modalities, which is a requirement for evaluating PET/CT alignment. Our goal is to evaluate several existing phantom designs to evaluate their utility for checking PET/CT alignment. The three phantoms investigated are a Gammex 464 phantom, a Triple-Line Source PET phantom, and a Hot Sphere PET phantom. The PET phantoms are unmodified the Gammex 464 phantom is modified to perform PET/CT alignment. The Gammex 464 phantom is typically used for routine quality assurance of CT scanners. Several CT parameters are determined with this phantom before and after modification. Then PET/CT alignment testing is performed using this modified CT phantom and the two other phantoms. Three methods have been used for analyzing the PET/CT images to measure the PET/CT alignment errors. The methods are the Manual method which calculates the alignment error from hand-drawn profiles, the Maximum-Pixel Value method which measures the error based on the pixel value of the objects in the PET/CT images, and the Curve-fitting method, which measures the alignment error by getting the best fit values for the object profiles. The Curve-fitting method also estimates the PET resolution from apparent size of objects in the phantoms. Our PET/CT alignment data and results suggest that the Maximum-Pixel Value method for the modified phantom with acrylic insert is a good choice for measuring the PET/CT alignment error, providing a reasonable balance between computational analysis effort and measurement precision.

Date

2005

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Kenneth L. Matthews II

Share

COinS