Results: Chronic psychosis was associated with grey matter deficitsbilaterally in the insular cortices, and in the superior frontal gyrus andleft caudate relative to the healthy group. No significant differenceswere detected between the first-episode and either the chronicpsychosis or healthy control groups. Grey matter deficits did notappear to be related to the severity of positive, negative, general or totalillness symptoms, the age of illness onset or medication exposure.Discussion: Grey matter deficits observed later in illness coursemay not be present to the same extent at the time of first episodeof psychosis. However, this would warrant further investigationusing small volume correction applied to examine regions stronglyimplicated in the pathophysiology of schizophrenia includingfrontal cortical regions. Our findings are in line with those of theprevious literature in identifying reduced grey matter in schizo-phrenia in frontal and temporal cortical regions.

doi:10.1016/j.schres.2010.02.860

Poster 100MAPPING RELIABILITY OF MULTICENTER MRI: CORTICALTHICKNESS AND VOXEL-BASED MORPHOMETRY

Hugo G. Schnack1, Neeltje E.M. van Haren1, Rachel M. Brouwer1,G. Caroline M. van Baal1, Marco Picchioni2,3, Matthias Weisbrod4,Heinrich Sauer5, Tyrone D. Cannon6, Matti Huttunen7,Claude Lepage8, D. Louis Collins8, Alan Evans8, Robin M. Murray3,Rene S. Kahn1, Hilleke E. Hulshoff Pol11Rudolf Magnus Institute of Neuroscience, Department of Psychiatry,University Medical Centre Utrecht Utecht, Utrecht, Netherlands; 2StAndrews Academic Centre, King's College London, Institute ofPsychiatry Northampton United Kingdom; 3Institute of PsychiatryLondon United Kingdom; 4Department of Psychiatry, University ofHeidelberg Heidelberg Germany; 5Department of Psychiatry, Universityof Jena Jena Germany; 6Departments of Psychology and Psychiatry andBiobehavioral Sciences, University of California Los Angeles, California,USA; 7Department of Mental Health and Alcohol Research, NationalHealth Institute Helsinki Finland; 8McConnell Brain Imaging Center,Montreal Neurological Institute, McGill University Montreal Canada

Background: Multicenter structural MRI studies can have greaterstatistical power to detect disease effects or heritabilities of braintissue than single-center studies. However, across-center differ-ences in contrast sensitivity, spatial uniformity, etc., may lead totissue classification or image registration differences that couldreduce or wholly offset the enhanced statistical power of multi-center data. Prior work has validated volumetric multicenter MRI,but robust methods for assessing reliability and power of multisiteanalyses with voxel-based morphometry (VBM) and corticalthickness measurement (CORT) are not yet available.Methods: We developed quantitative methods to investigate thereproducibility of VBM and CORT to detect group differences andestimate heritability when MRI scans from different scannersrunning different acquisition protocols in a multicenter setup areincluded. The method produces brain maps displaying informationsuch as lowest detectable effect size (or heritability) and effectivenumber of subjects in the multicenter study. We applied the methodto a five-site multicenter calibration study using scanners from fourdifferent manufacturers, running different acquisition protocols.Results: The reliability maps showed an overall good comparabilitybetween the sites, providing a reasonable gain in sensitivity in mostparts of the brain. In large parts of the cerebrum and cortex scanpooling improved heritability estimates, with 'effective-N' valuesupto the theoretical maximum. For some areas, 'optimal-pool' mapsindicated that leaving out a site would give better results. The

reliability maps also reveal which brain regions are in any casedifficult to measure reliably (e.g., around the thalamus).Discussion: These tools will facilitate the design and analysis ofmultisite VBM and CORT studies for detecting group differences andestimating heritability. The comparability between scans from thefive sites of the analyzed multicenter tudy is good, especially in thecortical areas. Combining MRI data from these different sites willresult in improved statistical power, allowing smaller effects to bediscovered.

doi:10.1016/j.schres.2010.02.861

Poster 101DIFFUSION TENSOR IMAGING OF THE CINGULUM BUNDLE INFIRST EPISODE SCHIZOPHRENIA

Jason S. Schneiderman1, Thomas J. Whitford1, Paula E. Pelavin1,Douglas P. Terry1, Tali Swisher1, Raquelle I. Mesholam-Gately2,Larry J. Seidman2, Jill M. Goldstein2, Robert W. McCarley3,Marek Kubicki1, Martha E. Shenton1

1Brigham and Women's Hospital, Harvard Medical School Boston, MA,USA; 2Beth Israel Deaconess Medical Center, Harvard Medical SchoolBoston, MA, USA; 3VA Boston Healthcare System Brockton, MA, USA

Background: The cingulum bundle is the major white matter tractconnecting the cortex and limbic system, and has been observed tobe abnormal in patients with chronic schizophrenia.Methods: The cingulum bundles of 19 patients with first episodeschizophrenia (FESZ) and 20 controls (NC) were examined usingdiffusion tensor imaging on a 3T MRI scanner.Results: Fractional anisotropy (FA) abnormalities were not signifi-cant between FESZ andNC.However, bothAxial andRadial diffusivitywere abnormal on both the left (FA: t(37)=0.763, p=0.450, Cohen'sd=0.244; Axial: t(37)=4.245, p<0.001, Cohen's d=1.360; Radial:t(37)=2.235, p=0.032, Cohen's d=0.716), and the right side (FA:t(37)=1.626, p=0.113, Cohen's d=0.521; Axial: t(37)=4.164,p<0.001, Cohen's d=1.334; Radial: t(37)=2.762, p=0.009, Cohen'sd=1.885) in FESZ compared to NC.Discussion: These results suggest that both axial and radialdiffusivity contribute to the change in overall white matter healthand organization measured by FA. Results further suggest that theunderlying pathology in FESZ in the cingulum bundle involvesabnormalities in both axon integrity and myelination. This isconsistent with abnormal myelin development that results inpartial degeneration of the underlying axons.

doi:10.1016/j.schres.2010.02.862

Poster 102CEREBRAL WHITE MATTER CHANGES ON COMBINEDSTRUCTURAL MRI AND DIFFUSION TENSOR IMAGINGIN FIRST EPISODE SCHIZOPHRENIA

KangSim1,W.Y. Chan1,M.Y. Chia1, G.L. Yang2, Y.Y. Sitoh3,W.L.Nowinski21Institute of Mental Health Singapore, Singapore, Singapore; 2SingaporeBiomedical Imaging Consortium, Agency for Science, Technology andResearch (ASTAR) Singapore, Singapore, Singapore; 3National Neu-roscience Institute Singapore, Singapore, Singapore

Background: Previous studies have revealed volumetric abnormal-ities of white matter in patients with schizophrenia but thecorresponding white matter dysconnectivities are less studied intandem. The aim of this study is to examine white matter integrity

Abstracts 461

in the region of white matter volume deficit in patients with first-episode schizophrenia (FES).

Methods: A cross-sectional, case-control design was adopted andwe used empirically-defined region of known white matter volumedeficit to interrogate diffusion tensors. The participants included103 subjects comprising of 39 patients with FES and 64 age-, sex-,and handedness-matched healthy controls. The neurocognitivedomains assessed included intelligence, attention, executive func-tioning, verbal and spatial working memory. The main outcomeswere gray and white-matter partial volumes, fractional anisotropy,trace and geometric diffusion indices.Results: Structural voxel-wise analyses revealed that patients withfirst episode schizophrenia had lower gray matter volumes inbilateral hippocampi (P<.01) and lower white matter volume inthe right temporal-occipital region (P<.005) corresponding to theinferior longitudinal fasciculus. Further analyses of diffusionanisotropy in the right temporal-occipital region revealed lowerplanar anisotropy, cp, and higher linear anisotropy, cl (P=0.012) inpatients with first episode schizophrenia. However, no differenceswere found for fractional anisotropy and trace in the implicatedwhite matter region between the two groups. Patients performedpoorer in digit span, spatial working memory and executivefunctioning, compared to healthy controls.Discussion: To the best of our knowledge, this is the first study toemploy geometric diffusion measures in interrogating the nature ofdiffusion tensor in schizophrenia. We confirmed previous findings ofwhite matter volume deficit in the region of inferior longitudinalfasciculus. The presence of changes in geometric diffusion indices inthe implicated white matter region suggests that pathophysiologicalprocesses which underlie cerebral white matter volume reductionmay not be reflected by changes in fractional anisotropy. Furtherresearch is needed to better understand the nature of these whitematter changes and its progression in schizophrenia over time.

doi:10.1016/j.schres.2010.02.863

Poster 103ALTERATIONS IN WHITE MATTER MICROSTRUCTURE ASSOCIATEDWITH THE ONSET OF PSYCHOSIS

Francesco Carletti1, James Woolley1, Matthew Broome1,4, ElviraBramon1, Louise Johns1, Paolo Fusar Poli1, Lucia Valmaggia1, VincentGiampietro3, Gareth J. Barker2, Philip K. McGuire11Psychosis Clinical Academic Group, Institute of Psychiatry, King'sHealth Partners, King's College London, London, London, UnitedKingdom; 2Centre for Neuroimaging Sciences, Department of ClinicalNeuroscience, Institute of Psychiatry, King's College London, London,London, United Kingdom; 3Brain Image Analysis Unit, Institute ofPsychiatry, King's College London, London, London, United Kingdom;4Psychosis Research Group, Health Science Research Institute, MedicalSchool, University of Warwick, Coventry, Warwick, United Kingdom

Background: The At Risk Mental State (ARMS) refers to a group ofsymptoms associated with a very high risk of developing psychosis:25–33% of ARMS subjects become psychotic within 1–2 years.Diffusion Tensor Imaging studies indicate that schizophrenia isassociated with reduced integrity in the frontal, cingulate, parietaland temporal white matter, but the extent to which vulnerability topsychosis is associated with abnormalities of structural connectivityis unclear. We addressed this issue through cross-sectional andlongitudinal comparisons of diffusion tensor imaging data acquiredfrom subjects with an ARMS.Methods: Thirty–one individuals meeting PACE criteria for the ARMSand 34 healthy volunteers were studied using a 1.5T MRI scanner.

Diffusion weighted data were acquired using a multi-slice, wholebrain, peripherally gated echo-planar imaging (EPI) sequence, opti-mized for precise measurement of the diffusion tensor in the brain. Ateach slice location 7 imageswithout diffusionweighting gradients and64 diffusion-weighted images (b-value=1300smm−2)with gradientdirections uniformly distributed in space were acquired. ARMS andfirst episode subjects were scanned at clinical presentation. Twenty-two of the ARMS subjects and 14 controls were re-scanned usingidentical methods after a mean interval of 28months. Five of these 22ARMS subjects had developed schizophrenia, defined according toDSM-IV criteria, in the 2 years subsequent to scanning. Between-groupdifferences in fractional anisotropy were evaluated longitudinally andcross-sectionally using a non-parametric voxel based analysis. Resultswere localised using the ICBM-DTI-81 white matter labels atlas andvalues of fractional anisotropy (FA), mean and radial diffusivity andthree eigenvalues extracted.Results:Baseline analyses:In relation to healthy volunteers, the ARMS group showed a reductionof FA in the left external capsule, in the posterior limb and in theretrolenticular portion of the left internal capsule, the left superior andposterior corona radiata, the left posterior thalamic radiation, and theleft superior longitudinal fasciculus. Within the ARMS group, thesubgroupwho later developed psychosis had reduced FA values in theright anterior corona radiata compared to those who did not.Longitudinal analyses:Within the ARMS group, in subjects who later developed psychosiscompared to those who did not, there was a longitudinal reduction inFA in the genu and body of the corpus callosum, the superior andposterior corona radiata and superior longitudinal fascicle bilaterally,the left anterior corona radiata, and the right external capsule,posterior limb of the right internal capsule and the right superiorfronto-occipital fasciculus.Discussion: These results suggest that abnormalities in whitematter tracts linking the frontal, parietal and temporal lobes areevident in people at high risk of psychosis. Within this group, thelater onset of psychosis was particularly associated with reducedintegrity in the major pathways between cortical areas, andbetween cortical and subcortical regions.

doi:10.1016/j.schres.2010.02.864

Poster 104A FOLLOW-UP MRI STUDY OF THE SUPERIOR TEMPORALSUBREGIONS IN SCHIZOTYPAL DISORDER ANDFIRST-EPISODE SCHIZOPHRENIA

Tsutomu Takahashi1,3, Michio Suzuki1,3, Shi-Yu Zhou4,Ryoichiro Tanino1, Kazue Nakamura1, Yasuhiro Kawasaki1,3,Hikaru Seto2, Masayoshi Kurachi51Department of Neuropsychiatry, University of Toyama Toyama Japan;2Department of Radiology, University of Toyama Toyama Japan;3CREST, JST, Tokyo, Japan; 4Department of Psychiatry and MedicalPsychology, Dalian Medical University Dalian China; 5Departments ofPsychiatric Early Intervention, University of Toyama Toyama Japan

Background: While longitudinal magnetic resonance imaging(MRI) studies have demonstrated progressive gray matter reductionof the superior temporal gyrus (STG) during the early phases ofschizophrenia, it remains unknown whether patients with schizo-typal features exhibit similar STG changes.Methods: In this study, longitudinal MRI data were obtained from 18patients with first-episode schizophrenia, 13 patients with schizotypaldisorder, and 20 healthy controls. The volumes of the STG and its

Abstracts462