Congrats to Tayebeh for her MSc thesis defense!
Title: Uncertainty in Probabilistic Image Registration
A challenging but important problem in image registration is evaluating the performance of a registration algorithm. Some methods opted to estimate registration accuracy in clinical data, where the Ground Truth (GT) is typically unknown. For example, some methods used uncertainty measures as a surrogate for quantitative registration error. Other methods collected training data with GT warps and trained machine learning algorithms to infer registration error or to improve the registration results for novel data.
In this thesis, we first examine existing uncertainty definitions and propose a new spatially-based uncertainty measure, ExpErMAP. Subsequently, we use ExpErMAP, that is superior to existing measures in the expected idealized behaviour, correlation with noise and error, ability to detect pathology, and computational complexity, to evaluate and improve registration results in a learning framework. Our method consists of a training and a testing stage.
M.Sc. Examining Committee:
Dr. Ghassan Hamarneh, Senior Supervisor
Dr. Hao (Richard) Zhang, Supervisor
Dr. Torsten Möller, Examiner
Dr. Mark Drew, Chair