Optimizing and advancing multiparametric magnetic resonance imaging for biologically-guided radiotherapy
Loading...
Date
2019-12-18
Authors
Bergen, Robert
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Multi-parametric MRI studies of the prostate often include anatomical images, as well as functional images such as perfusion-weighted images (PWI) and diffusion-weighted images (DWI). These functional images give insights into tissue micro-environment which helps physicians further differentiate between healthy and cancerous tissue. MRI can also potentially measure tissue hypoxia, the lack of oxygen within the tissue, which can introduce resistance to radiotherapy treatment and negatively affect treatment outcomes. Ideally, these data sets would be used to characterize a dominant prostate lesion, to which a radiation dose could be escalated during radiotherapy. However, incorporating DWI, PWI and oxygenation measurements into treatment planning is not routine, because the imaging requirements for MRI-guided radiotherapy are stricter than diagnostic imaging
requirements. For DWI, image distortion may be a significant source of error, and
therefore must be minimized. PWI imaging relies heavily on T1 mapping, but conventional
T1 mapping methods can be very inaccurate and affect the localization or characterization of the dominant lesion. Finally, oxygenation measurements in the prostate using advanced imaging techniques like quantitative susceptibility mapping (QSM) requires validation. Tissue oxygen level dependent (TOLD) imaging, another method that is sensitive to oxygenation, also requires additional corrections due to its sensitivity to temperature.
To improve upon conventional multi-parametric MRI, correction methods were implemented to reduce image distortion in DWI and to reduce uncertainties in T1 mapping for PWI. The correction methods were implemented both in phantom and in vivo and compared to conventional techniques. The feasibility of oxygenation measurements, using both QSM and TOLD, was also tested in phantom and in vivo, and temperature
measurements were acquired so that a correction could be applied to the TOLD data. Finally, the optimized and conventional imaging methods were compared in terms of prostate lesion localization and characterization. Simulations were then performed to investigate the effects on a prostate treatment plan. Significant differences were found between the optimized and conventional DWI and PWI imaging techniques, and the feasibility of MR oxygenation measurements was demonstrated. It was shown that the improved multi-parametric MRI acquisitions had significant impacts on lesion localization and characterization, which could potentially have significant effects on treatment planning.
Description
Keywords
MRI, Diffusion, Perfusion, Radiotherapy, Susceptibility mapping