PDF version of this entire document

next up previous
Next: Introduction

Non-Rigid Registration Assessment Without Ground Truth

R. S. Schestowitza, C. J. Twininga, T. F. Cootesa, V. S. Petrovica, W. R. Crumb, and C. J. TayloraE-mail addresses for correspondence: Roy.Schestowitz@stud.man.ac.uk; Carole.Twining, V.Petrovic, Tim.Cootes, Chris.Taylor@manchester.ac.uk; Bill.crum@cs.ucl.ac.uk

aImaging Science and Biomedical Engineering, Stopford Building,
University of Manchester, Oxford Road, Manchester M13 9PT, UK.
bCentre for Medical Image Computing, Department of Computer Science,
University College London,Gower Street, London WC1E 6BT, UK.


We compare two methods for assessing the performance of groupwise non-rigid registration algorithms. The first approach, which has been described previously, utilizes a measure of overlap between ground-truth anatomical labels. The second, which is new, exploits the fact that, given a set of non-rigidly registered images, a generative statistical model of appearance can be constructed. We observe that the quality of this model depends on the quality of the registration, and define measures of model specificity and generalisation - based on comparing synthetic images sampled from the model, with those in the original image set - that can be used to assess model/registration quality. We show that both approaches detect the loss of registration accuracy as the alignment of a set of correctly registered MR images of the brain is progressively perturbed. We compare the sensitivities of the two approaches and show that, as well as requiring no ground truth, specificity provides the most sensitive measure of misregistration. Finally, we use specificity and generalisation to compare three NRR algorithms.

next up previous
Next: Introduction
Roy Schestowitz 2007-03-11