 |
 |
Dear SPM-list,
I have a dataset for which I’m not sure whether everything is alright:
One thing that strikes me is that there seems to be “too much” significance (I know, sounds strange :-)
In a high-level contrast (i.e. comparing rather similar conditions, not a resting baseline contrast or alike), I find beta-values in the lateral-prefrontal cortex of 10-15. Usually I have less than 2-3 in those areas for such contrasts.
Peak voxels have t-value of up to 15 (not that unusual) and p(FWE!) is often <
.00001
One thing I noticed is that the resels are very large. Here’ what the SPM Results graphics window provides:
Voxel size: 2x2x2 mm (resel = 2368.70 voxels)
Volume = 1159408 = 144926 voxels = 56.6 resels
FWHM = 27.5 26.6 25.9mm; 13.7 13.3 13.0 voxels
I smoothed using 8mm kernel.
Preprocessing and analysis was pretty “textbook standard” ( N = 17):
Realign & Unwarp; Normalization (via Unified Segmentation); Smoothing (8 mm FWHM Gaussian Kernel); First level stats; First level contrasts; Second level stats (one-sample t-tests).
The data were analysed twice so far, once with SPM8 (some years ago) and now re-analysed using SPM12. Both times same results. Overall, the activated areas are meaningful (i.e., the expected areas are activated).
Besides the high resel count, this is my only dataset from a certain scanning site – I never scanned there again (GE 3T scanner). For a similar dataset (N=16) from another scanner I get a resel size of 413 voxels (same processing pipeline
and parameters), i.e. roughly only 1/6th the size.
Are these above values plausible – but I just have never happen to observe them before? Or are they indicative of anything wrong in the data, some mistake in the processing, user error, etc?
I tried to look at the images, and to me they look fine (by visual inspection).
Many thanks &
Kind Regards,
Andre
_______________________________________________________
Dr André J. Szameitat
Reader in Psychology
Co-Director Centre for Cognition and Neuroimaging (CCNI)
T +44(0)18952
67387 | E [log in to unmask]
Brunel University London
Dept of Life Sciences
Psychology
Gaskel Building, Room GASK263
Office hours: Wed 11.30-12.30, Thu 13.30-14.30
Brunel University London, Uxbridge, UB8 3PH, United Kingdom
Connect with the University on Facebook, Twitter, LinkedIn, Google+, Instagram
_______________________________________________________