dc.rights.license | CC-BY-NC-ND | |
dc.contributor.advisor | Externe beoordelaar - External assesor, | |
dc.contributor.author | Bochove, Aart van | |
dc.date.accessioned | 2022-02-22T00:00:31Z | |
dc.date.available | 2022-02-22T00:00:31Z | |
dc.date.issued | 2022 | |
dc.identifier.uri | https://studenttheses.uu.nl/handle/20.500.12932/508 | |
dc.description.abstract | Introduction
Cardiorespiratory motion causes large positional uncertainty in radiotherapy of the thoracic and abdominal regions. Respiratory-induced uncertainty reduces when treating in the time-weighted average/mid-position (midP) anatomy. This thesis investigates combining a coarse resolution T2w respiratory-correlated 4D-MRI with a high-resolution end-exhale triggered T2w MRI scan (MVXD), to yield a high-resolution midP. Furthermore, including cardiac triggering during MVXD acquisition is proposed. The quality and feasibility of higher-resolution midP and cardiorespiratory triggered MVXD are assessed.
Methods & materials
The respiratory-triggered MVXD and a 4D-MRI were acquired for thirteen patients and two healthy volunteers. MidP images, constructed by warping the 4D-MRI phases to the MVXD scan using deformable image registration, were compared to midPs created conventionally by only using the 4D-MRI. The consistency of the deformable vector fields (DVFs) was assessed using the distance discordance metric (DDM).
Five healthy volunteers were scanned with a cardiorespiratory-triggered MVXD, where the 1D-navigator was placed on the left ventricle (LV) instead of the liver-lung (LL-)interface. 2D cine-MRI images were used to determine residual motion. Three observers assessed the MVXD image quality.
Results
When including the MVXD in midP calculations, MidP resolution increased nine-fold, yielding sharper images compared to 4D-only midP-MRI. DDM values increased (8% DDM values >2 mm to 17% >2 mm), signaling reduced DVF consistency, mostly in areas with pulsatile flow (heart, blood vessels) and ghosting (skin), as expected.
Using LV-based cardiorespiratory triggering, halved the LL- and LV-motion compared to no triggering. According to observers, changing LL-triggering to LV-triggering caused improved heart sharpness in the MVXD images, while LL-interface sharpness decreased slightly. Average scan time increased from 5:46 to 5:51.
Conclusion & Discussion.
Involving MVXD in midP calculations increases midP image quality, but DVF consistency slightly reduces, probably due to contrast differences between the scans. Using cardiorespiratory triggering increases MVXD quality around the heart.
Including the MVXD scan in midP calculation is feasible and useful, and applying cardiorespiratory triggering can increase MVXD quality. | |
dc.description.sponsorship | Utrecht University | |
dc.language.iso | EN | |
dc.subject | This thesis describes combining a T2w respiratory-correlated 4D-MRI with a high-resolution T2w end-exhale triggered MRI scan to yield a high-resolution mid-position image. Furthermore, including cardiac triggering with the high-resolution MRI scan is proposed, to increase the image quality. Both methods are evaluated. | |
dc.title | Improved motion-compensated MRI and image processing for mid-position lung radiotherapy | |
dc.type.content | Master Thesis | |
dc.rights.accessrights | Open Access | |
dc.subject.keywords | Lung, radiotherapy, MRI, mid-position, cardiorespiratory triggering, image processing | |
dc.subject.courseuu | Medical Imaging | |
dc.thesis.id | 2379 | |