Use this URL to cite or link to this record in EThOS:
Title: Estimation and reliability of myocardial blood flow with dynamic PET
Author: Saillant, Antoine
ISNI:       0000 0004 7971 5788
Awarding Body: University of Oxford
Current Institution: University of Oxford
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Dynamic Positron Emission Tomography (PET) enables the estimation of Myocardial Blood Flow (MBF), by modeling the temporal evolution of a radio-labeled pharmaceutical over multiple PET images acquired over time. Despite its advantages, the absolute quantification of MBF is not used routinely in the clinic due to a lack of clinical standardization. MBF estimates can be affected by patient motion, among other factors, which creates displacements between different PET frames and leads to errors in MBF calculation. Automated motion correction techniques are difficult to perform, due to the rapid change of the tracer spatial distribution throughout the PET images series, and the low resolution of PET image series. Furthermore, it is challenging to quantitatively evaluate the success of motion correction techniques, and their impact on MBF reliability is not fully understood. In this thesis, we developed a framework for estimating MBF along with the uncertainty on the estimate. The framework relies on Bayes theory, which represents the kinetic parameters as a probability distribution, from which uncertainty measures can be extracted. If the uncertainty extracted is high the parameter studied is considered to have high variability - or low confidence - and vice versa. A Bayesian model inference technique, called Variational Bayes, was chosen due to its relatively inexpensive computations compared to other Bayesian techniques. The reliability of parameter and uncertainty estimates were first evaluated through simulations and compared with another Bayesian model inference technique considered the "gold standard". It was found that the uncertainty metric was estimated robustly across multiple noise levels, which was necessary as motion can be seen as an additional source of noise. The uncertainty metric was then used to assess motion correction, performed manually by experienced observers. The uncertainty, calculated before and after motion correction, reduced significantly in moderate and high motion cases, demonstrating the usefulness of the method. The uncertainty was subsequently used to evaluate a motion correction algorithm, showing comparable performance to manual realignment provided by experienced users, with regards to the uncertainty metric. Finally, we investigated different situations where the use of the uncertainty metric may convey clinical information, besides assessing motion correction.
Supervisor: Chappell, Michael ; Jenkinson, Mark Sponsor: Siemens Healthineers
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available