Rdiger Brhl1, Albrecht Ihlenfeld1, Semiha Aydin1, Penelope A. Gowland2, Gareth John Barker3, Tomas Paus4, Jrgen Finsterbusch5, Simone Khn6, Jrgen Gallinat7, Arun Bokde8, Michael Smolka9, Frauke Nees10, Jean-Luc Martinot11, Gunter Schumann3, the IMAGEN Consortium3, Bernd Ittermann1
1PTB, Berlin, Germany; 2University of Nottingham, Nottingham, United Kingdom; 3Institute of Psychiatry, Kings College London, London, United Kingdom; 4Rotman Research Institute, University of Toronto, Toronto, Ontario, Canada; 5University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 6Max-Planck-Institut fr Bildungsforschung, Berlin, Germany; 7Psychiatry and Psychotherapy, Charit Universittsmedizin, Berlin, Germany; 8Institute of Neuroscience, Trinity College Dublin, Dublin, Ireland; 9Department of Psychology, Technische Universitt Dresden, Dresden, Germany; 10Central Institute of Mental Health, Mannheim, Germany; 11University Paris Descartes, Paris, France
The multicentre study IMAGEN scanned more than 2000 at 8 sites in Europe using scanners of 4 different manufactures. Freesurfer's volumetric results show significant differences between the sites and especially the scanner models. Data of phantoms and 2 traveling volunteers are used to explain these differences leading to the necessity of characterizing the gradient induced distortion field locally. Data of 2 new precision geometry phantoms will provide the information to finally decide whether the differences are artificial or sample generated.