Problem with two-pass normalisation
The standard aa normalisation routine ([http://imaging.mrc-cbu.cam.ac.uk/imaging/AutomaticAnalysisVersions#head-20f22d73e91b461dcf6bf69e792269bdce21ecf4 Automatic Analysis normalisation with segment]) may fail in certain cases. The example below illustrates it: at the bottom, the normalised brain on the left is smaller and has not been reoriented correctly.
attachment:diagnostic_aamod_norm_noss.jpg
In order to allow monitoring of such occurrences, please mail to imagerstech@mrc-cbu.cam.ac.uk if you notice this problem.
Possible contributing problems were (1) that the subject had a small brain to start with (2) that they were very obese, with alot of signal from fat around the scalp and neck (3) we did not reposition and reorient the images (eg with SPM display option) prior to normalisation.
Our solution: For this participant, we used the SPM GUI 'normalize' procedure instead of the aamod_norm_noss (combined 'Segment and normalise') and write modules. Using Check registration (in SPM) the normalisation appeared OK