 |
 |
Dear Karel, > is ther a way to mask the images before the statistical analysis? I'm > particularly interested in one region of the brain, but the whole brain > is scanned. After the statistical analysis there are more blobs than the > region I'm interested in. I want to remove these blobs by masking my > smoothed and normalised images before the analysis. Is this possible > and statistically correct? You can restrict your analysis to a portion of the brain without having to do this, by using the Small Volume Correction after statistical analysis. This allows you to confine examination (and appropriate statistical correction) of your results to just a specified volume of brain. The Small Volume can be defined either as a sphere/box around a specified coordinate (if that's enough to localise the region you're interested in), or as an image (e.g. ROI created outside SPM). There's been some previous correspondence on on the list (if you search) about how to create an ROI image outside SPM for this purpose. Best wishes, Geraint --- Geraint Rees | Institute of Cognitive Neuroscience [log in to unmask] | University College London +44-(0)20-7670-5496 (work) | Alexandra House +44-(0)20-7813-2835 (fax) | 17 Queen Square www.klab.caltech.edu/~geraint | London WC1N 3BG