Time

Prog #

Peng Wang¹, Ruben Gur¹, Ragini Verma¹

¹Univ. of Pennsylvnia, Philadelphia, Pennsylvania, USA

Diffusion tensor imaging (DTI) provides rich information about brain tissue structure [1], especially white matter and has therefore gained attention in studying pathology of several diseases, such as schizophrenia, in a group-based analysis. In such research, it is important to accurately identify abnormal brain tissue that can classify patients and controls into two groups, for diagnosis, treatment and prognosis. However, conventional statistical methods are difficult to apply to DTI data due to its high dimensionality and nonlinearity. Most of the existing methods resort to a statistical t-test based on FA or kernel PCA features from tensors. However, the t-test usually assumes a single Gaussian distribution for each group, and a multivariate hypothesis testing is usually constrained by limited number of available samples and high-dimensional data. To address these issues, we present a novel group analysis framework in this paper. Rooted in the pattern classification theory, our method directly estimates the overlap between different groups using Bayes error rate. Unlike t-test, our method does not assume global Gaussian distribution for each group, but only considers data distribution to be locally smooth, thus facilitating separation of different groups when the data could have highly nonlinear structure, which is the case in DTI data. This framework also has the capability of combining multiple measurements that are extracted from tensor data, thus being able to handle multivariate testing. In this paper, the method is first used to identify abnormal brain regions that distinguish the patient group from control group, and is further validated by classifying patients from controls based on these regions. Experiments on 36 controls and 34 patients with schizophrenia show encouraging results.

Adam W. Anderson¹, Mevhibe N. Tumuklu², Lori R. Arlinghaus¹, Tricia A. Thornton-Wells¹, Xin Hong¹, Herbert Y. Meltzer¹

¹Vanderbilt University, Nashville, Tennessee, USA; ²Gaziosmanpasa University Faculty of Medicine, Tokat, Turkey

Diffusion tensor imaging was used to measure the variations in white matter structure related to single nucleotide polymorphisms (SNPs) in the histamine H1 receptor gene. Statistical parametric maps reveal regions of white matter where anisotropy depends strongly on SNP genotype. Hence, DTI can be used to map the effects of genotype on brain structure, thereby providing an 'intermediate phenotype' that may help to improve understanding of the relationships between genes and behavior.

David Matthew Carpenter¹, Cheuk Tang¹, Joseph I. Friedman¹, ², Hof R. Patrick, Daniel G. Stewart¹, Monte S. Buchsbaum¹, Philip D. Harvey¹, Jack M. Gorman³, Kenneth L. Davis

¹Mount Sinai School of Medicine, New York, New York, USA; ²Pilgrim Psychiatric Center, West Brentwood, New York, USA; ³Mount Comprehensive NeuroScience, Inc., White Plains, New York, USA

Here we  implemented a fiber tracking algorithm and used it to investigate the temporal nature of white matter abnormalities in schizophrenia. Quantitative DTI tractography was used to quantify the fractional anisotropy (FA) of the forceps minor and the pyramidal tracts bilaterally.  The tractography approach showed a significant decline in FA that was correlated with the duration of illness in the forceps minor.  This FA decline in patients is beyond that accounted for by normal aging.

Babak A. Ardekani¹, ², Jay Nierenberg¹, ², Debra D’Angelo¹, Matthew J. Hoptman¹, ²

¹Nathan Kline Institute, Orangeburg, USA; ²New York University School of Medicine, New York, USA

The main objective of this study was to determine if using FA and ADC images concurrently in linear discriminant analysis (LDA) results in increased classification accuracy as compared to using FA or ADC alone.   We scanned 42 patients with schizophrenia and 42 normal controls subjects.   A two-dimensional LDA was performed following feature extraction using singular value decomposition.   Specificity, sensitivity, and overall accuracy were respectively 85.7%, 71.4%, and 78.6% for ADC images alone, 88.1%, 73.8%, and 81.0% for FA images alone, and 90%, 76.2%, and 83.3% when both ADC and FA were used for classification.  The results suggest that using multiple brain imaging measures may improve discrimination accuracy.

Su Lui¹, Xiaoqi Huang¹, Wei Deng¹, Hehan Tang¹, Lijun Jiang¹, Dongming Li¹, Tijiang Zhang¹, Xiuli Li¹, Tao Li¹, Qiyong Gong¹, ²

¹West China Hospital of Sichuan University, Chendu, People's Republic of China; ²University of Liverpool, Liverpool, UK

Optimized voxel based morphometry (VBM) in conjunction with the analysis of the resting state functional brain connectivity was applied to determine the relationship of the cerebral deficits with the symptoms in drug-naïve first episode schizophrenia (FSE).  21 drug-naïve FSE and 21 age, sex, height, weight, handedness and years of education matched controls were recruited, and were scanned using a volumetric 3D SPGR sequence and an EPI sequence on a 3T MR imaging system.  For first time, we characterized that the symptoms of FSE associated with both the regional structural deficits and the functional networks.

Vaibhav A. Diwadkar¹, ², Eric Murphy¹, Simon B. Eickhoff³, Matcheri S. Keshavan¹, ²

¹Wayne State University SOM, Detroit, Michigan, USA; ²University of Pittsburgh SOM, Pittsburgh, Pennsylvania, USA; ³Forschungszentrum Juelich, Juelich, Germany

Schizophrenia is characterized in part by hippocampal related pathology.  The hippocampus is the primary site in the brain for the initial consolidation of associative memories, and this process depends in part on NMDA receptor function.  Schizophrenia is hypothesized to involve NMDA receptor hypofunction; therefore associative learning is an ideal paradigm using which to assess hippocampal function in the illness.  We separately assessed changes in fMRI measured activity in the dentate gyrus and cornu ammonis sub-regions in the hippocampus during associative memory and demonstrate distinct patterns of pathological response in these sub-regions in schizophrenia.

Mark Greenwald¹, Dalal Khatib¹, Matcheri Keshavan¹, Jeffrey Stanley¹

¹Wayne State University, Detroit, Michigan, USA

High-field short TE single-voxel 1H-MRS is used to determine in heroin-dependent subjects whether glutamate (GLU) levels in anterior cingulate cortex (ACC) and thalamus vary during within-subject manipulation of methadone maintenance dose. Subjects are first stabilized on high-dose methadone (scan 1), followed by reduction to a low dose (scan 2).  All 6 completers in this ongoing study exhibit significantly lower GLU in ACC during high- vs. low-dose methadone. Heroin craving during high-dose methadone is positively correlated with change in ACC GLU during dose reduction. These findings are consistent with preclinical data and suggest new theoretical and therapeutic directions.

Ai-Ling Lin¹, David C. Glahn², Ahmad R. Hariri³, Douglas E E. Williamson²

¹University of Texas Health Science Center , San Antonio, Texas, USA; ²University of Texas Health Science Center, San Antonio, Texas, USA; ³University of Pittsburgh, Pittsburgh, Pennsylvania, USA

The purpose of this study was to investigate alterations in baseline cerebral blood flow (CBF) in adolescents at risk for alcohol use disorder (AUD). The baseline CBF was measured in both  low- and high-risk groups for AUD using pulsed ASL sequence. Compare to the low-risk group, the high-risk group shows higher CBF in bilateral amygala and ventromedial prefrontal cortex but less CBF in bilateral insula and occipital cortex. These preliminary findings suggest that adolescents at relatively high-risk for AUD exhibit altered patterns of resting CBF in distributed corticolimbic regions supporting emotional behaviors.

Anton Orlichenko¹, K Luan Phan², Huiling Peng¹, Emil F. Coccaro³, Konstantinos Arfanakis¹

¹Illinois Institute of Technology, Chicago, Illinois, USA; ²University of Michigan, Ann Arbor, Michigan, USA; ³University of Chicago, Chicago, Illinois, USA

Social anxiety disorder, also known as social phobia, refers to long-lasting anxiety, extreme distress and impaired ability to function, caused by perceived social threats. It has in the past been correlated with hyperactivation of the amygdala, and hypoactivation of ventral frontal areas believed to provide cognitive regulation of the limbic system. The aim of this study was to use Turboprop diffusion tensor imaging to examine differences in white matter microstructure between patients with social phobia and healthy controls.

Yann Gagnon¹, Tim Devito¹, Janet Hendry¹, N Gelman¹, N Rajakumar², P Williamson², R Nicolson², Dick Drost¹

¹University of Western Ontario, Lawson Health Research Institute, London, Canada; ²University of Western Ontario, London, Canada

The transverse relaxation time (T2) is a quantitative parameter of magnetic resonance imaging which is indicative of the molecular environment in brain tissue.  In this study, T2 times of the basal ganglia were evaluated in a group of Tourette Syndrome (TS) patients.  Images obtained at 3 Tesla using the GESFIDE technique were spatially normalized for statistical analysis.  TS patients were found to have increased T2 of the caudate nucleus when compared with controls.