Improved Methods for the Registration and Motion Correction of Brain Images

FMRIB Technical Report TR02MJ1
(A related paper has been accepted by Neuroimage)

Mark Jenkinson$^{1}$, Peter Bannister$^{1,2}$, Michael Brady$^{2}$ and Stephen Smith$^{1}$

1: Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB),
Department of Clinical Neurology, University of Oxford, John Radcliffe Hospital,
Headley Way, Headington, Oxford, UK
2: Medical Vision Laboratory, Department of Engineering Science,
University of Oxford, Oxford, UK
Corresponding author is Mark Jenkinson: mark@fmrib.ox.ac.uk

Abstract:

Linear registration and motion correction are important components of structural and functional brain image analysis. Most modern methods optimise some intensity-based cost function to determine the best registration. To date, little attention has been focused on the optimisation method itself, even though the success of most registration methods hinges on the quality of this optimisation. This paper examines the optimisation process in detail and demonstrates that the commonly used multi-resolution local optimisation methods can, and do, get trapped in local minima. To address this problem, two approaches are taken: (1) to apodize the cost function and (2) to employ a novel hybrid global-local optimisation method. This new optimisation method is specifically designed for registering whole brain images. It substantially reduces the likelihood of producing mis-registrations due to being trapped by local minima. The increased robustness of the method, compared to other commonly used methods, is demonstrated by a consistency test. In addition, the accuracy of the registration is demonstrated by a series of experiments with motion correction. These motion correction experiments also investigate how the results are affected by different cost functions and interpolation methods.