An MRI study of structural variations in schizophrenia using deformation field morphometry

Psychiatry Research: Neuroimaging 146 (2006) 171 – 177 www.elsevier.com/locate/psychresns

An MRI study of structural variations in schizophrenia using deformation field morphometry Uicheul Yoona, Jong-Min Leea,*, Jun Soo Kwonb, Hyun-Pil Kima, Yong-Wook Shinb, Tae Hyon Hab, In Young Kima, Kee-Hyun Changc, Sun I. Kima a

Department of Biomedical Engineering, Hanyang University, Sungdong P.O. Box 55, Seoul, 133-605, Korea b Department of Psychiatry, Seoul National University College of Medicine, Seoul, Korea c Department of Radiology, Seoul National University College of Medicine, Seoul, Korea Received 21 September 2005; received in revised form 6 December 2005; accepted 14 December 2005

Abstract Magnetic resonance imaging (MRI) has an important role in investigating the changes in brain structure that are associated with schizophrenia. In this study, MRI scans of patients diagnosed with schizophrenia (37 males; 19 females; 17–42 years of age) were compared with those of an age- and sex-matched group of normal subjects (37 males; 19 females; 18–40 years of age). Based on the images of the healthy control subjects, we constructed a representative average brain template. Automated image analysis techniques were used to measure differences in the regional nonlinear deformation fields between the two groups. A deformation field, which measures the spatial transformation to deform a template of brain anatomy to each individual data, was obtained as a three-dimensional displacement vector in each voxel. There was a significantly greater magnitude of the deformation fields in the superior frontal and parietal lobes as well as in the cingulate gyrus connecting both lobes of the patients with schizophrenia than in those of healthy controls, suggesting that these cerebral regions have a significantly higher structural variability in schizophrenia. D 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Schizophrenia; Magnetic resonance imaging; Average template; Deformation fields; Structural variations

1. Introduction Schizophrenia is a debilitating illness that affects almost 1% of the population. Although its etiology is not fully understood, schizophrenia is thought to have its origins in brain chemistry and/or cortical abnormalities. The morphological characteristics (Supprian et al., 1997; Okazaki, 1998; Pfefferbaum et al., 1999; Kova* Corresponding author. Tel.: +82 2 2220 0685; fax: +82 2 2296 5943. E-mail address: [email protected] (J.-M. Lee). URL: http://cna.hanyang.ac.kr (J.-M. Lee).

lev et al., 2003) of the cortical changes include enlargement of the lateral ventricles (Andreasen et al., 1990), greater variability of gray matter (Narr et al., 2005), and alterations in the volume of certain lobes or substructures (see McCarley et al., 1999; Shenton et al., 2001, for more extensive reviews). These morphological abnormalities could be explained by disruptions between cerebral substructures (Buchsbaum, 1990; Breier et al., 1992; Weinberger, 1996; Saunders et al., 1998) or problems occurring during neurological development (Feinberg, 1982; Crow, 1997, 1998). With advances in magnetic resonance imaging (MRI) and computational analysis of neuroanatomy,

0925-4927/$ - see front matter D 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2005.12.005

172

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177

it is now possible to compare the in vivo morphology of the whole brain across multiple subjects. When a subject’s brain MRI is warped (deformed) to become identical to a template, a deformation field is generated, which consists of the vector (distance and direction) that each point in the subject’s MR image had to follow to achieve a successful warp to a template (a process referred to as nonlinear registration). Ashburner et al. (1998) and Gaser et al. (1999) used deformation fields to analyze global and local differences between two groups. Their automated methods were able to detect morphological changes. A number of deformation-based methods to detect morphological changes with multivariate statistics have been developed in recent years (Thompson et al., 1997; Gaser et al., 1999; Carmack et al., 2004). These new methods are highly automated, free from intra- and inter-operator variability, and do not depend on an a priori determination of a region of interest (ROI). So far, research on schizophrenia using structural MRI has been limited to detecting volumetric changes or limited analyses of dominant shape abnormalities. Reviews by McCarley et al. (1999) and Shenton et al. (2001) showed that there were several brain regions that demonstrated neither positive nor negative volumetric differences between patients with schizophrenia and healthy controls from structural MRI, in spite of apparent abnormalities in functional analysis (Wible et al., 1995) and cell density (Selemon et al., 1998). They suggested that this discrepancy might be related to insufficient sensitivity of MRI methods to small volumetric changes that were statistically significant. On the other hand, Park et al. (2004) investigated the spatial variability of specific ROIs in first-episode schizophrenic patients using probability maps. They reported most ROIs of patients with schizophrenia showed a significantly lower spatial overlap than in controls, suggesting a greater heterogeneity in the spatial distribution of ROIs. Although this finding was based on voxelwise comparisons of local areas by the use of statistical probability difference maps, it would be necessary to present objective and quantitative measures representing the differences in spatial distribution in specific regions and the whole brain. In this article, we assessed the structural variations of patients with schizophrenia over the entire brain in terms of deformation field, rather than volumetric or shape changes. We statistically analyzed magnitude fields of deformation to the template through the nonlinear spatial normalization.

2. Methods 2.1. Subjects The study followed guidelines for the use of human subjects established by the institutional review board, and all subjects participating in the study gave their written, informed consent. A group of right-handed patients with schizophrenia (37 males; 19 females; 17–42 years of age) was recruited from the Seoul National University Hospital, Seoul, Korea. A healthy control group of subjects (37 males; 19 females; 18 40 years of age) was selected by advertising on the Internet and was matched with the affected group for sex, age, handedness, and parental socioeconomic status (SES). Parental SES and personal SES were assessed using the five-point Hollingshead scale (1 = highest, 5 = lowest) (Hollingshead and Redlich, 1958). The demographic characteristics of patients and controls are summarized in Table 1. There were no significant differences between the affected and control groups in mean age and parental SES. Variables such as years of education, estimated intelligence quotient and personal SES were not matched, and the expected differences in these variables between the groups were significant. Each patient with schizophrenia was interviewed using the Structured Clinical Interview for DSM-IV (American Psychiatric Association, 1994), and each met criteria for schizophrenia. In the schizophrenia group, the severity of each patient’s symptoms was rated on the Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987). The mean PANSS score across all patients with schizophrenia was 74.4 F 17.7. None of the

Table 1 Demographic characteristics for the patients with schizophrenia and healthy control subjects Characteristic

Age (years) Sex (M/F) Education (years) Estimated IQa Parental SESb SESb PANSSc total score a

Schizophrenia (n = 56)

Control (n = 56)

Mean

S.D.

Mean

S.D.

27.2 37/19 14.3

5.6

25.4 37/19 15.6

5.1

1.7

0.08

2.2

3.2

b0.01

101.5 3.0 3.2 74.4

11.5 0.8 0.7 17.7

116.0 3.0 2.8

10.7 0.7 0.6

5.8 0.4 2.9

b0.01 0.7 b0.01

2.2

Student t-test t

P

IQ, intelligence quotient. SES, socioeconomic status, assessed using the Hollingshead scale. Higher scores indicate lower status. c PANSS, Positive and Negative Syndrome Scale. b

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177

173

schizophrenia or normal control subjects had a history of neurological disease, significant medical illnesses or substance abuse, and none of the controls had a history of a DSM-IV axis I disorder. 2.2. Image acquisition and preprocessing Magnetic resonance images were acquired on a 1.5 T scanner (SIGNA, GE Medical Systems, Milwaukee, WI, USA) using a T1-weighted 3D-spoiled gradient recalled echo pulse sequence with the following parameters: 1.5-mm sagittal slices; echo time (TE) = 5.5 ms; repetition time (TR) = 14.4 ms; number of excitations = 1; rotation angle = 208; field of view (FOV) = 21  21 cm; matrix = 256  256. All images were resampled to an isocubic voxel (0.82  0.82  0.82 mm3) and realigned so that the anterior–posterior axis of the brain was parallel to the intercommissural line, and the other two axes were aligned along the interhemispheric fissure (Kim et al., 2001). The data were prepared using our previously described methods (Yoon et al., 2003). 2.3. Template construction The Automated Image Registration (AIR) program (version 5.2.5) was used for the linear and nonlinear registrations (Woods et al., 1998a,b). Linear registration is a process that allows for global spatial normalization. It is undertaken to eliminate global differences from all images, such as initial position and overall size. This alignment of all images allows for and is a prerequisite of voxel-wise comparison between subjects. Nonlinear registration refers to the process of warping (deforming) a subject’s MR image to become identical to a template. The basic procedure of template construction was derived and modified from the bAIR Make AtlasQ pipeline (Rex et al., 2003). The process consisted of three steps in which an initial template was made followed by linear and nonlinear atlases (Fig. 1). First, to reduce variations of global position and scale, all initial MR images were linearly registered to a randomly chosen initial brain within a group, which was visually inspected to screen out severe irregularity. The registered images were averaged and non-brain structures were masked out, thus forming the initial template. To prevent the registration procedure from being severely biased by using an individual image as a registration target, all intensity-normalized initial MR images were then linearly registered to the initial template, averaged and masked again, thereby yielding the linear atlas. Finally, all MR images were nonlinearly registered

Fig. 1. Three-step procedure for producing an average brain atlas as a common spatial coordinate. First, the average template was constructed as a bias-free registration target. Then, linear registration was performed based on the average template for global scaling and linear alignment. Finally, nonlinear fifth-order polynomial warping was completed to eliminate natural variations caused by the nonrigidity of brains in vivo.

and warped onto the linear atlas. This nonlinear registration makes sure that the subject’s MR image becomes identical to the template, and produces a deformation field through the process. By examining this deformation field, we could measure the distance and direction of change in each voxel for achieving the registration. 2.4. Deformation field analysis The hypothesis tested in this analysis was that there would be significant differences in the deformation distances if one group of subjects had a higher spatial variability than another group. We quantified the displacement in terms of the regional and nonlinear deformation distance of each voxel in the template of the healthy control group. Extracting the nonlinear regional deformation field using regional spatial normalization takes place following linear registration of all individual brain images to the average template of healthy controls, thereby ensuring that regional spatial normalization captured only the nonlinear and regional differences. Nonlinear deformation parameters were then generated with AIR’s polynomial warping transform, and the regional deformation fields in three-dimensional vector form were extracted with an in-house program that used internal functions of the AIR package. In each subject, these vector fields were converted to their magnitude fields to map the distance moved by a voxel in a template image. The magnitude fields in this phase were not continuous due to the loss of dimensional information, so they were filtered using a Gaussian kernel at a full-width half-maximum (FWHM) of 8 mm. Finally, a mapping program

174

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177

(SPM2, Wellcome Department of Imaging Neuroscience, University College, London, UK) was used for the voxel-wise statistical analysis (two-sample t-test) of the magnitude fields. Note that the program was not used for voxel- or deformation-based morphometry but for voxel-wise statistical tests; i.e., our analysis using the mapping program was not related to its built-in brain templates. 3. Results There were two types of spatial normalization such as inter-group and intra-group registration. In this study, the average template generated from only the healthy controls were used for the target of spatial normalization. Therefore, the deformation of the patients with schizophrenia to the average template represents intergroup registration and that of the healthy control population stands for intra-group registration. The magnitude fields of deformation generated from each case were averaged respectively to examine the pattern of structural variability. It ranged from 0 to 6.1 mm and showed significant differences in the magnitude fields between groups (Fig. 2a and b). The structural variations of the patients with schizophrenia were significantly greater than those of the healthy controls as shown by the bright yellow regions in Fig. 2. Note

that the greatest deformation was close to the boundaries of the cerebral cortex, which were highly variable in the sulcal/gyral pattern. The voxel-wise statistical analysis of the magnitude fields showed that the superior frontal and parietal lobes as well as the cingulate gyrus connecting both lobes in patients with schizophrenia were significantly greater than those of healthy controls ( P b 0.001, uncorrected), but there was no significant difference in any other region of the brain (Fig. 2c). 4. Discussion 4.1. Clinical implication of structural variations in schizophrenia Previous studies of patients with schizophrenia have reported changes in cerebral volume (Shenton et al., 2001). There were reductions of volume in the frontal and temporal lobes as well as whole brain of subjects either experiencing their first psychotic episode or those previously treated (Gur et al., 1998). Mathalon et al. (2001) have also reported significant reduction of the volume of gray matter in the right frontal lobe. In the frontal lobe, a clear association between significant abnormalities in functional analysis (Wible et al., 1995) and cell density (Selemon et al., 1998) has

Fig. 2. (a) The mean magnitude fields of the deformation of healthy control (HC-AT_ HC) and (b) of schizophrenia groups (SZ-AT_ HC) were obtained with respect to the average template generated from the healthy control (AT_ HC), respectively. Note that HC-AT_ HC can be viewed as intra-group registration and SZ-AT_ HC as inter-group registration. The mean magnitude fields of the deformation fields derived from the patients with schizophrenia show larger displacement across its lobular area than those of the healthy controls. (c) Statistical analysis between the magnitude fields. The color-coded area denotes where the magnitude fields of deformation from the patients with schizophrenia are larger than those of the healthy controls. The color bar represents the t-values of each voxel. The wide area across the superior frontal and parietal lobes, as well as the superior cingulate gyrus, is significantly larger in patients with schizophrenia than in healthy controls. However, no significant area was found in the opposite case.

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177

been established. Structural MRI has demonstrated alterations in the volume of the frontal lobe or subcortical structures (McCarley et al., 1999; Shenton et al., 2001). However, the changes were not statistically significant since MRI was not sensitive enough to detect these small alterations. As for the parietal lobe, Niznikiewicz et al. (2000) demonstrated correlations between the inferior parietal lobule and the neuroanatomically connected cortical regions of both the prefrontal (superior and inferior frontal gyrus and orbital gyrus) and temporal regions of the cortex, thereby supporting the idea that the associative cortex and interrelated brain regions might be affected in schizophrenia. Our results showed that the patients with schizophrenia generally had greater magnitude fields of the deformation across the superior frontal and parietal lobes, which were connected by the cingulate gyrus, than healthy control subjects. Because the measurement for the statistical analysis was based on the regional difference of deformation to the template obtained after removing global differences across subjects and template, it could be said that the patients with schizophrenia had a greater structural and spatial variability in their superior frontal and parietal lobes as well as the cingulate gyrus. Enlarged lateral and third ventricles might affect the structural variability in these adjacent regions (Lawrie and Abukmeil, 1998). However, we believe our findings support the existence of a spatial perturbation; i.e., in schizophrenia, there are associated structural variations across the interrelated sub-regions of the brain that specifically affect the neural pathways connecting the superior frontal and parietal lobes, and the superior portion of the cingulate gyrus. These results could be confirmed in a structural or functional connectivity study with diffusion tensor imaging. Although our finding does not fully correspond with a previous ROI-based study (Park et al., 2004), it could support that there is greater structural variability in schizophrenia. Therefore, we suggest that it is worth doing this kind of analysis for whole brain spatial variations since it can shed light on other aspects of structural changes. 4.2. Methodological considerations Because an inappropriately chosen target might invalidate the registration process, various techniques have been devised to choose the optimal target for registration (Woods et al., 1999; Kochunov et al., 2001, 2002). In this study, a simple technique that produced an average template was adopted for the

175

initial registration target. Use of this technique, instead of the choice of an optimized individual target brain, leads to a reduction in the sharpness of the template, because individual structural variability was averaged and smoothed. However, the average template significantly reduced the influence of a single brain on the subsequent registration process so that the final template was less biased and more generally represented the whole population. The whole procedure was integrated in a highly automated and reproducible way, and thus, comparisons could be made between many samples with minimum intervention of the operator. Unlike the deformation- or voxel-based morphometric techniques, our method focused on the distribution of structural variations in the whole brain rather than a localized or regional change in shape or volume (Gaser et al., 1999; Job et al., 2002). Another method for the analysis of brain variability has been developed by Narr et al. (2001) and Thompson et al. (2001), but their work mainly focused on the variability of the cortical surface, not the whole brain. 5. Conclusion In the present study, we found that there was a greater distance of deformation to the template in the frontal and parietal lobes as well as the cingulate gyrus of patients with schizophrenia than healthy controls. These findings suggest that patients with schizophrenia have a greater structural variation in those regions. Acknowledgments This work was supported by grant R11-2002-103 from the Korean Science and Engineering Foundation. References American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, fourth ed. American Psychiatric Publishing, Washington, DC. Andreasen, N.C., Ehrhardt, J.C., Swayze II, V.W., Alliger, R.J., Yuh, W.T., Cohen, G., Ziebell, S., 1990. Magnetic resonance imaging of the brain in schizophrenia. The pathophysiologic significance of structural abnormalities. Archives of General Psychiatry 47, 35 – 44. Ashburner, J., Hutton, C., Frackowiak, R., Johnsrude, I., Price, C., Friston, K., 1998. Identifying global anatomical differences: deformation-based morphometry. Human Brain Mapping 6, 348 – 357. Breier, A., Buchanan, R.W., Elkashef, A., Munson, R.C., Kirkpatrick, B., Gellad, F., 1992. Brain morphology and schizophrenia. A magnetic resonance imaging study of limbic, prefrontal

176

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177

cortex, and caudate structures. Archives of General Psychiatry 49, 921 – 926. Buchsbaum, M.S., 1990. The frontal lobes, basal ganglia, and temporal lobes as sites for schizophrenia. Schizophrenia Bulletin 16, 379 – 389. Carmack, P.S., Spence, J., Gunst, R.F., Schucany, W.R., Woodward, W.A., Haley, R.W., 2004. Improved agreement between Talairach and MNI coordinate spaces in deep brain regions. NeuroImage 22, 367 – 371. Crow, T.J., 1997. Temporolimbic or transcallosal connections: where is the primary lesion in schizophrenia and what is its nature? Schizophrenia Bulletin 23, 521 – 523. Crow, T.J., 1998. Schizophrenia as a transcallosal misconnection syndrome. Schizophrenia Research 30, 111 – 114. Feinberg, I., 1982. Schizophrenia: caused by a fault in programmed synaptic elimination during adolescence? Journal of Psychiatric Research 17, 319 – 334. Gaser, C., Volz, H.P., Kiebel, S., Riehemann, S., Sauer, H., 1999. Detecting structural changes in whole brain based on nonlinear deformations—application to schizophrenia research. NeuroImage 10, 107 – 113. Gur, R.E., Cowell, P., Turetsky, B.I., Gallacher, F., Cannon, T., Bilker, W., Gur, R.C., 1998. A follow-up magnetic resonance imaging study of schizophrenia. Relationship of neuroanatomical changes to clinical and neurobehavioral measures. Archives of General Psychiatry 55, 145 – 152. Hollingshead, A.B., Redlich, F.C., 1958. Social Class and Mental Illness: A Community Study. Wiley, New York. Job, D.E., Whalley, H.C., McConnell, S., Glabus, M., Johnstone, E.C., Lawrie, S.M., 2002. Structural gray matter differences between first-episode schizophrenics and normal controls using voxel-based morphometry. NeuroImage 17, 880 – 889. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophrenia Bulletin 13, 261 – 276. Kim, J.J., Lee, M.C., Kim, J., Kim, I.Y., Kim, S.I., Han, M.H., Chang, K.H., Kwon, J.S., 2001. Grey matter abnormalities in obsessive– compulsive disorder: statistical parametric mapping of segmented magnetic resonance images. British Journal of Psychiatry 179, 330 – 334. Kochunov, P., Lancaster, J.L., Thompson, P., Woods, R., Mazziotta, J., Hardies, J., Fox, P., 2001. Regional spatial normalization: toward an optimal target. Journal of Computer Assisted Tomography 25, 805 – 816. Kochunov, P., Lancaster, J.L., Thompson, P., Toga, A.W., Brewer, P., Hardies, J., Fox, P., 2002. An optimized individual target brain in the Talairach coordinate system. NeuroImage 17, 922 – 927. Kovalev, V.A., Petrou, M., Suckling, J., 2003. Detection of structural differences between the brains of schizophrenic patients and controls. Psychiatry Research: Neuroimaging 124, 177 – 189. Lawrie, S.M., Abukmeil, S.S., 1998. Brain abnormality in schizophrenia. A systematic and quantitative review of volumetric magnetic resonance imaging studies. British Journal of Psychiatry 172, 110 – 120. Mathalon, D.H., Sullivan, E.V., Lim, K.O., Pfefferbaum, A., 2001. Progressive brain volume changes and the clinical course of schizophrenia in men: a longitudinal magnetic resonance imaging study. Archives of General Psychiatry 58, 148 – 157. McCarley, R.W., Wible, C.G., Frumin, M., Hirayasu, Y., Levitt, J.J., Fischer, I.A., Shenton, M.E., 1999. MRI anatomy of schizophrenia. Biological Psychiatry 45, 1099 – 1119.

Narr, K., Thompson, P., Sharma, T., Moussai, J., Zoumalan, C., Rayman, J., Toga, A., 2001. Three-dimensional mapping of gyral shape and cortical surface asymmetries in schizophrenia: gender effects. American Journal of Psychiatry 158, 244 – 255. Narr, K.L., Bilder, R.M., Toga, A.W., Woods, R.P., Rex, D.E., Szeszko, P.R., Robinson, D., Sevy, S., Gunduz-Bruce, H., Wang, Y.P., DeLuca, H., Thompson, P.M., 2005. Mapping cortical thickness and gray matter concentration in first episode schizophrenia. Cerebral Cortex 15, 708 – 719. Niznikiewicz, M., Donnino, R., McCarley, R.W., Nestor, P.G., Iosifescu, D.V., O’Donnell, B., Levitt, J., Shenton, M.E., 2000. Abnormal angular gyrus asymmetry in schizophrenia. American Journal of Psychiatry: Neuroimaging 157, 428 – 437. Okazaki, Y., 1998. Morphological brain imaging studies on major psychoses. Psychiatry and Clinical Neuroscience 52, S215 – S218. Supplement. Park, H.J., Levitt, J., Shenton, M.E., Salisbury, D.F., Kubicki, M., Kikinis, R., Jolesz, F.A., McCarley, R.W., 2004. An MRI study of spatial probability brain map differences between first-episode schizophrenia and normal controls. NeuroImage 22, 1231 – 1246. Pfefferbaum, A., Sullivan, E.V., Hedehus, M., Moseley, M., Lim, K.O., 1999. Brain gray and white matter transverse relaxation time in schizophrenia. Psychiatry Research 91, 93 – 100. Rex, D.E., Ma, J.Q., Toga, A.W., 2003. The LONI pipeline processing environment. NeuroImage 19, 1033 – 1048. Saunders, R.C., Kolachana, B.S., Bachevalier, J., Weinberger, D.R., 1998. Neonatal lesions of the medial temporal lobe disrupt prefrontal cortical regulation of striatal dopamine. Nature 393, 169 – 171. Selemon, L.D., Rajkowska, G., Goldman-Rakic, P.S., 1998. Elevated neuronal density in prefrontal area 46 in brains from schizophrenic patients: application of a three-dimensional, stereologic counting method. Journal of Comparative Neurology 392, 402 – 412. Shenton, M.E., Dickey, C.C., Frumin, M., McCarley, R.W., 2001. A review of MRI findings in schizophrenia. Schizophrenia Research 49, 1 – 52. Supprian, T., Hofmann, E., Warmuth-Metz, M., Franzek, E., Becker, T., 1997. MRI T2 relaxation times of brain regions in schizophrenic patients and control subjects. Psychiatry Research: Neuroimaging 75, 173 – 182. Thompson, P.M., MacDonald, D., Mega, M.S., Holmes, C.J., Evans, A.C., Toga, A.W., 1997. Detection and mapping of abnormal brain structure with a probabilistic atlas of cortical surfaces. Journal of Computer Assisted Tomography 21, 567 – 581. Thompson, P.M., Mega, M.S., Woods, R.P., Zoumalan, C.I., Lindshield, C.J., Blanton, R.E., Moussai, J., Holmes, C.J., Cummings, J.L., Toga, A.W., 2001. Cortical change in Alzheimer’s disease detected with a disease-specific population-based brain atlas. Cerebral Cortex 11, 1 – 16. Weinberger, D.R., 1996. On the plausibility of bthe neurodevelopmental hypothesisQ of schizophrenia. Neuropsychopharmacology 14, 1S – 11S. Wible, C.G., Shenton, M.E., Hokama, H., Kikinis, R., Jolesz, F.A., Metcalf, D., McCarley, R.W., 1995. Prefrontal cortex and schizophrenia. A quantitative magnetic resonance imaging study. Archives of General Psychiatry 52, 279 – 288. Woods, R.P., Grafton, S.T., Holmes, C.J., Cherry, S.R., Mazziotta, J.C., 1998a. Automated image registration: I. General methods and intrasubject, intramodality validation. Journal of Computer Assisted Tomography 22, 139 – 152.

U. Yoon et al. / Psychiatry Research: Neuroimaging 146 (2006) 171–177 Woods, R.P., Grafton, S.T., Watson, J.D., Sicotte, N.L., Mazziotta, J.C., 1998b. Automated image registration: II. Intersubject validation of linear and nonlinear models. Journal of Computer Assisted Tomography 22, 153 – 165. Woods, R.P., Dapretto, M., Sicotte, N.L., Toga, A.W., Mazziotta, J.C., 1999. Creation and use of a Talairach-compatible atlas for accurate, automated, nonlinear intersubject registration, and

177

analysis of functional imaging data. Human Brain Mapping 8, 73 – 79. Yoon, U.C., Lee, J.M., Kim, J.J., Lee, S.M., Kim, I.Y., Kwon, J.S., Kim, S.I., 2003. Modified magnetic resonance image based parcellation method for cerebral cortex using successive fuzzy clustering and boundary detection. Annals of Biomedical Engineering 31, 441 – 447.