> 10 mm accuracy for spatial normalization sounds a lot for images that start
> with resolutions around 1 mm. Is SPM the best in terms of sccuracy of
> registration or are there better tools? Do you know of any literature
> comparing various methods in this regard?
The best references I know are by Hellier et al:
@article{ hellier01retrospective,
author = "P. Hellier and C. Barillot and I. Corouge and B. Gibaud and G.
Le Goualher and L. Collins and A. Evans and G. Malandain and N. Ayache",
title = "Retrospective Evaluation of Inter-subject Brain Registration",
journal = "Lecture Notes in Computer Science",
volume = "2208",
pages = "258--??",
year = "2001",
url = "citeseer.nj.nec.com/hellier01retrospective.html" }
@InProceedings{Hellier02a,
Author = {Hellier, P. and Ashburner, J. and Corouge, I. and Barillot, C.
and Friston, K.J.}, Title = {Inter subject registration of functional and
anatomical data using SPM}, BookTitle = {Medical Image Computing and
Computer Assisted Intervention 2002, MICCAI'02, LNCS 2489}, Pages =
{590--587},
Address = {Tokyo},
Month = {September},
Year = {2002}
}
You may be able to get these from:
[log in to unmask]" target="_blank">http:[log in to unmask]
Average distance between registered sulci and corresponding sulci of the
reference subject. The mean is computed for all the subjects and all the
sulci (central, superior frontal and sylvian):
9.9mm ANIMAL
10.3mm Demons
11.5mm Rigid-registration
10.7mm Proportional squaring of Talairach
10.8mm Pierre Hellier's method
8.7mm SPM99
Skull-stripped brains were used, and no brain-masking. Other than that,
the SPM99 default settings were used for the registration.
If manually defined landmarks are also used, then more accurate registration
can be achieved (especially near the landmarks).
See Hellier & Barillot.IEEE TMI 22(2):217-227 (2003)
Best regards,
-John
--
Dr John Ashburner.
Functional Imaging Lab., 12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420 http://www.fil.ion.ucl.ac.uk/~john
|