Uncertainties in prostate targeting during radiotherapy: assessment, implications and applications of statistical methods of process control

Ngie Min Ung

    Research output: ThesisDoctoral Thesis

    351 Downloads (Pure)

    Abstract

    [Truncated abstract] In the last 15 years, a great deal of effort has been expended on developing and investigating various aspects of image-guided radiotherapy (IGRT). IGRT of the prostate gland is of particular interest due to the highly mobile nature of the organ and its proximity to radiosensitive organs. This thesis sets out to quantify some uncertainties that may impact IGRT processes for the specific case of prostate targeting using fiducial markers. Methods inspired by Statistical Process Control (SPC) for improving existing techniques and setup correction strategies in IGRT of the prostate are also demonstrated. A number of uncertainties that may affect the accuracy of fiducial tracking using an electronic portal imaging device (EPID) were investigated. The impacts of beamcollimation device errors, inter- and intra-observer variability and the subtended angles for stereoscopic registration using the radiation field-edge detection method were first quantified experimentally via a phantom study. This study was extended to evaluate the impact of errors of the beam-collimation device on treatment plans for intensitymodulated radiotherapy (IMRT) of the prostate. Results showed that errors in beamcollimation devices affect the accuracy of localization measurements as well as the dose distribution of IMRT plans for prostate treatment. However, assuming good practices in machine quality assurance are applied, the clinical impacts of these geometric collimation errors are dwarfed by the potential magnitude of the patient-related displacment of the prostate.
    Original languageEnglish
    QualificationDoctor of Philosophy
    Publication statusUnpublished - 2012

    Fingerprint Dive into the research topics of 'Uncertainties in prostate targeting during radiotherapy: assessment, implications and applications of statistical methods of process control'. Together they form a unique fingerprint.

    Cite this