Abstract #0893

On the Comparability of Volumetric Brain Data in the Multicentric IMAGEN Study

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.