I'm preparing a paper regarding evaluation of brain image registrations. From one paragraph of the Introduction:

"...Guido Gerig developed the Valmet software tool (http://www.ia.unc.edu/public/valmet/) and Gary Christensen and collaborators are creating the Non-rigid Image Registration Evaluation Project (NIREP) (http://www.nirep.org). The Windows-based Valmet was in 2001 the first publicly available software tool for measuring (as well as visualizing) the differences between corresponding image segmentations. It uses several algorithms to compare segmentations: overlap ratio, Hausdorff distance, surface distance, and probabilistic overlap. The NIREP pro ject “has been started to develop, establish, maintain, and endorse a standardized set of relevant benchmarks and metrics for performance evaluation of nonrigid image registration algorithms.” The initial phase of the project will include 16 manually labeled brain images (32 labeled regions in 8 men and 8 women) and four evaluation metrics: (1) relative overlap (equivalent to the "union overlap" defined in the Materials and methods section), (2) variance of the registered intensity images for an image population, (3) inverse consistency error between a forward and reverse transformation between two images, and (4) transitivity (how well all the pairwise registrations of the image population satisfy the transitivity property). One of the goals of NIREP is to incorporate the metrics used by Valmet (9). Another significant goal is to make all findings available in publications and on their website."

Cheers,
@rno

arno klein
http://www.binarybottle.com