Original Article
Although most functional and structural neuroimaging studies have reported that cognitive deficits in patients with schizophrenia are associated with prefrontal abnormalities, their findings vary considerably. The aim of this study is to find evidence of functional abnormalities in the prefrontal cortex of schizophrenia patients on the basis of exact topographical parcellation. Magnetic resonance (MR) images were obtained in 12 patients with schizophrenia and 12 age-and sex-matched normal volunteers, and parceled into 8 frontal subregions using topographic landmarks. [15O]H2O PET scans were obtained during the visual working memory task and the control task, and the activities of the parceled frontal subregions were counted on the registered PET images. In the comparison of the subregional functional activities, the normal healthy group showed a tendency to activate the right rostral anterior cingulate during the working memory task. Comparatively, the patient group showed reduced activation of the right orbitofrontal cortex and the right caudal anterior cingulate. The results suggest that the functional changes observed in schizophrenia may result from the abnormal activation of the neural system for the purpose of performing working memory tasks. In particular, reduced ventromedial prefrontal activities seem to play an important role in such cognitive impairment in schizophrenia.
Correspondence: Jae-Jin Kim, MD, PhD, Associate Professor Department of Psychiatry and Diagnostic Radiology, Yonsei University College of Medicine Severance Mental Health Hospital, 696-6 Tanbul-ong, Gwangju, Gyeonggi-do 464-100, Korea
Tel: +82-31-760-9402, Fax: +82-31-761-7582, E-mail: jaejkim@yonsei.ac.kr
In the past two decades, functional activation studies have been increasingly used to explore the pathophysiology of schizophrenia. Among the studies performed so far, the most popular approach was to use cognitive tasks specific to the frontal lobe functions. For example, there are a number of studies showing reduced activity in the prefrontal regions during certain tasks, such as the Wisconsin card sorting test1, Tower of London Test2, and a verbal fluency task3. However, since such tasks basically require various executive functions including working memory, it is difficult to conclude which specific cognitive function relates to the prefrontal abnormalities in schizophrenia. Therefore, a cognitive task which specifically activates working memory only would be more useful in functional activation studies designed to explore the prefrontal abnormalities in schizophrenia. It has been proposed that working memory, which is the process of holding and
manipulating information online in the brain, consists of three main components, namely the central executive, visual sketch pad, and phonological loop4. In recent functional imaging studies, working memory, especially the central executive component, has been consistently shown to be closely related to prefrontal activities5,6,7. Working memory is important not only for the memory system itself, but also for the thinking process8. Impairments in working memory are viewed as constituting an important factor in the various cognitive abnormalities associated with schizophrenia, and are also thought to play a fundamental role in various clinical symptoms including formal thought disorder9. Recently, there have been a number of studies which demonstrated that abnormalities in prefrontal activation occur during the use of working memory in patients with schizophrenia, although the results are somewhat conflicting. Reduced PET or functional MR activities were found in the right dorsolateral prefrontal areas during visual working memory10,11 and auditory working memory operations12. However, another auditory working memory study, using a task requiring the recall of verbally given words and sounds, showed that patients with schizophrenia had decreased activities in the lateral prefrontal areas, but not in the dorsolateral prefrontal area13. In addition, there was also a study reporting an opposite result, in that patients with schizophrenia did not show decreases in activation in the prefrontal areas, but instead increases in the left dorsolateral prefrontal area14.
As mentioned above, despite the existence of a considerable number of reports showing a tendency toward hypofrontality during working memory tasks, the findings of the frontal subregions are inconsistent. The reason for this may be due to differences in the subject groups, arising from the heterogeneity of the schizophrenic patients involved in these studies. The variety of paradigms used for the activation study may be another reason. There are two experimental paradigms that can be used for addressing working memory, the delayed matching task and the n-back task; the former generally emphasizes the information holding function of the prefrontal area during a certain period of time15, whereas the latter includes information holding as well as the dynamic renewal of information through the prefrontal central executive8,16. Moreover, differences in the measurement methodologies employed constitute another important factor which needs to be considered. In particular, this aspect is more critical in approaches using regions of interest (ROIs). It is inevitable that many brain regions defined as large ROIs in previous studies simultaneously include activated and suppressed subregions17, and thus do not function homogenously. The findings of a particular study, therefore, depend on how the ROIs were defined. For these reasons, recent functional imaging studies have chosen the voxel-based approach, which is based on the linear proportional scaling system18. This has a significant benefit in that there is no need for artificial delineation of the ROIs, but has the disadvantage of causing the interindividual and interhemispheric variabilities to disappear. That is, the normalization of the results to a standard brain pattern, which is accomplished by performing linear and nonlinear transformation of the voxel-based analysis, forces the individuality of the functional organization to be effaced19,20.
Consequently, a new approach maintaining the individuality without artificial delineation is needed. This can be accomplished by topography-based delineation using sulcal landmarks on the MR images and measuring the activities on the coregistered PET images. Detailed topography-based parcellation methods have been reported in a series of studies21,22,23, and subregional volumetric abnormalities were demonstrated using such methods24.
This study was designed to investigate prefrontal functional abnormalities during the activation of working memory in patients with schizophrenia on the basis of topographically demarcated ROIs. We obtained MR images of patients with schizophrenia and normal healthy volunteers, subdivided the frontal lobe using the exact sulcal landmarks, and observed a difference in subregional PET activation during the working memory tasks between the two groups.
Materials and Methods
Subjects
The subjects were 12 patients with schizophrenia and 12 age- and sex-matched healthy volunteers. Each group contained 6 men and 6 women, all of whom were right-handed. The patients fulfilled the DSM-IV criteria for schizophrenia, as diagnosed by using the Structured Clinical Interview for DSM-IV (SCID-IV)25. The exclusion criteria for the study were any history of neurological or significant medical illnesses and any past history of substance abuse. The patient group had a mean age of 26.2 years (SD=4.0), a mean number of years of education of 14.7 (SD=1.6), a mean parental socioeconomic status26 of 2.9 (SD=1.1), and a mean duration of illness of 2.8 years (SD=3.2) at the time the PET scan was performed. Symptom severity was assessed with the Positive and Negative Syndrome Scale27; the patients' mean scores were 14.1 (SD=6.6) for the positive symptoms, 17.6 (SD=6.8) for the negative symptoms, 31.3 (SD=6.0) for the general symptoms, and 62.9 (SD=10.7) overall. All of the patients were taking one or two atypical neuroleptics, such as risperidone, olanzapine and clozapine. The healthy volunteers were recruited through a newspaper and internet advertisement, and were excluded if any history of psychiatric or neurological disorder was present. They had a mean age of 25.6 (SD=4.8), a mean number of years of education of 15.3 (SD=1.6), and a mean parental socioeconomic status of 3.2 (SD=1.0), with these values not being significantly different from those of the patient group. All of the subjects were given a full outline of the study and written informed consent was obtained.
Brain MRI and the Frontal Lobe Parcellation
Three-dimensional T1-weighted spoiled gradient echo MR images were acquired on a 1.5 Tesla GE SIGNA Scanner (GE Medical Systems, Milwaukee, USA). The imaging parameters were as follows: 1.5mm sagittal slices, echo time=5.5ms, repetition time=14.4ms, number of excitations=1, rotation angle=20 degrees, field of view=210x210mm, and a matrix of 256x256 was used. The MR images were processed using an image-processing software package, ANALYZE (version 3.0, Mayo Foundation, USA). The images were resampled to 1.0mm3 iso-voxels, reoriented to the conventional position, and spatially realigned, so that the anterior-posterior axis of the brain was aligned parallel to the inter-commissural line, and the other two axes were aligned along the inter-hemispheric fissure. The data sets were then filtered using anisotropic diffusion methods to improve the signal to noise ratio. In order to extract those parts of the images corresponding to the brain, tissues exterior to the brain were removed from the images by the semi-automated region growing method. By employing the fuzzy C-means algorithm28, the extracted brain images were segmented into three regions, namely the gray matter, white matter and cerebrospinal fluid. The intracranial volume was calculated by adding the subtotal volume of the three components. The frontal lobe was parcellated into 8 ROIs, namely the superior frontal gyrus, the middle frontal gyrus, the inferior frontal gyrus, the orbitofrontal cortex, the rostral anterior cingulate, the caudal anterior cingulate, the supplementary motor area and the precentral gyrus. The boundaries of each subregion were obtained from previous reports by a team in which one of the authors of the present study was included21, and a detailed description of the parcellation method can be found in a previous report by one of the authors24.
Brain PET and Cognitive Activation
PET imaging data acquisition
Scans were obtained using an ECAT EXACT 47 scanner (Siemens-CTI, Knoxville, TN, USA). Before the first injection of the tracer, a 7-min transmission scan was performed for the purpose of attenuation correction. Emission scans during the performance of the cognitive tasks started after an intravenous bolus injection of 30-50 mCi of
[15O]H2O in 5-7 ml saline and continued for 100 seconds, being recorded in the form of twenty 5-s frames. The injections were repeated at intervals of about 15 minutes. The acquired data were reconstructed in a 128x128x47 matrix with a pixel size of 2.1x2.1x3.4mm by means of a filtered back-projection algorithm employing a Shepp-Logan filter with a cut-off frequency of 0.3 cycles/pixel. Based on the time-activity curves, only the twelve frames, reflecting the 60 seconds after peak activity was reached, were summed and used to construct the static cerebral blood flow images.
Cognitive tasks
In order to avoid evoking a confounding factor by poor performance, all of the subjects were trained at an explanatory session before the real PET scan was performed, during which they had to read a set of instructions describing the rules of the task, and practice the procedure described in the written instructions until they were able to perform the performance correctly. The subjects performed six tasks, of which only two were used for this study's analysis. The first task reflected simple focused attention, and was used as a control task. In this task, a circle, a rectangle and a triangle were shown randomly and the subjects were told to respond whenever the circle was presented. The second task was an experimental task with a 2-back working memory condition, in which the subjects were asked to respond whenever the circle was presented and the penultimate stimulus was also the circle. Figure 1 shows the actual stimuli and the target position. Each task consisted of 80 stimuli items and the stimuli were projected for 0.5 seconds each at 1.5-second intervals. Each task had 28 targets among 80 stimuli items, and the frequency as well as the distribution of the targets were matched. The mouse-clicking responses of the subjects were automatically stored and used for the calculation of the adequacy of the performance. The adequacy of the performance was measured by evaluating the correct answer rate, according to the equation {(total number of stimuli -error response) /total number of stimuli}x100%. The errors included missing targets and mistakenly responding to non-target stimuli. The task stimuli were automatically projected at a distance of 40 cm from subjects' eyes by means of a 15 inch LCD monitor, and the subjects were asked to click the mouse with their index finger in order to respond.
Subregional Functional Measurement
Image preprocessing for the coregistration
The PET images were transferred to a personal computer, and converted to the ANALYZE software file format. In order to increase the signal to noise ratio, the images were smoothed by means of a 16mm FWHM Gaussian kernel included in Statistical Parametric Mapping (SPM) 99 (Institute of Neurology, University College of London, UK) implemented in Matlab (Mathworks Inc., USA). The smoothed images were coregistered on T1-weighted MR images of the same subject using ANALYZE, and the information concerning the frontal lobe parcellation was loaded onto the MR images. Figure 2 shows the traced frontal subregions and the registered PET image loaded on the MR image.
Measurement of the subregional activities
Because the total PET blood flow differed for each subject, preprocessing for the sake of calibration was done before calculating the blood flow of each separate subregion. After the mean image intensity of the total voxels was calculated, all voxels whose intensity was less than 12.5%of the calculated mean were eliminated, and the mean image intensity of the remaining voxels was recalculated. Since a standard blood flow rate of 50ml/min/dl was used for the recalculated mean image intensity, each voxel's blood flow rate was estimated by linear transformation. The adjusted mean ROI activity was calculated as the average of the estimated intensity values of all of the voxels for every parceled frontal subregion. Furthermore, the adjusted total ROI activity was calculated by multiplying the mean ROI activity by the ROI volume.
Statistical Analysis
The group differences in the demographic factors and the correct response rate for the cognitive task were analyzed using the Wilcoxon rank sum test. In the case of the mean activities of the frontal subregions, the Wilcoxon rank sum test was also used for the comparison between the two groups in each task condition, whereas the Wilcoxon signed rank test was used for the comparison between the two task conditions in each group, and repeated measures ANOVA was used for the between-group differences in the changing rates between the task conditions. The same analysis method was applied to the total activities of all frontal subregions.
Results
Task performance
In the focused attention condition, the correct response rate was 99.9% (SD=0.37) in the healthy group and 99.6% (SD=0.62) in the patient group; there was no significant difference between the two groups (p=0.18). In the working memory condition, the correct response rate was 99.2% (SD=1.0) in the healthy group and 97.9% (SD=3.5) in the patient group; there was no significant difference between the two groups (p=0.32).
Comparison of the mean subregional activities
The mean activities of the frontal subregions are shown in Table 1. During both the control and working memory conditions, the mean activities of all subregions were lower in the patient group than in the healthy group. The areas where the patient group showed lower activities than the healthy group were the right caudal anterior cingulate gyrus, both sides of the superior frontal gyrus, the middle frontal gyrus, the inferior frontal gyrus, and the rostral anterior cingulate gyrus during the control condition, and both sides of the middle frontal gyrus, the inferior frontal gyrus, the rostral anterior cingulate, the caudal anterior cingulate, and the orbitofrontal cortex during the working memory condition. In the comparison of the subregional functional activities, the normal healthy group showed a tendency for the right rostral anterior cingulate to be activated during the working memory condition, whereas the patient group did not. Comparatively, the patient group showed reduced activation of the right orbitofrontal cortex and the right caudal anterior cingulate, whereas the healthy group showed no changes. As shown in Figure 3, the activities of the right orbitofrontal cortex during the working memory condition were increased in the healthy group, but decreased in the patient group. The significance of this feature was confirmed by repeated measures ANOVA (p=0.006).
Comparison of the total subregional activities
The total activities of the frontal subregions are shown in Table 2. The total subregional activities during the control and working memory conditions were all lower in the patient group than in the healthy group. During the control condition, this tendency toward lower activity in the patient group was especially marked in the right supplementary motor area, both sides of the superior frontal gyrus, the precentral gyrus and the caudal anterior cingulate. In the case of the working memory condition, the right superior frontal gyrus, both sides of the middle frontal gyrus, the orbitofrontal cortex, the supplementary motor area and the caudal anterior cingulate showed lower total activities in the patient group, as compared with the healthy group. In addition, as in the case of the mean subregional activities during working memory, the normal healthy group showed significant activation only in the right rostral anterior cingulate, whereas the patient group showed significantly reduced activation in the right orbitofrontal cortex and the right caudal anterior cingulate.
Discussion
The key to the validity of this study's result lies in the exact topographical parcellation of the frontal cortex. The validity of the parcellation method used in this study was already proved in a previous study21. We used eight frontal subregions defined by topography-based parcellation as the areas of measurement of the functional activity. We assumed that using this method would substantially help us to overcome the limitations of the voxel-based analysis method that has been widely used in recent functional neuroimaging studies. The voxel-based analysis method spatially normalizes all brain images to a standard and, therefore, the individuality of functional organization is inevitably effaced. In the current study, however, we used the subjects' own MR image instead of the normalized brain and, in this way, we were able to investigate the functional changes during cognitive load for each frontal subregion, using boundaries based on anatomical landmarks. The core hypothesis which forms the basis of this new method is that the functional map of the neocortex is individually unique, and that this individuality is directly reflected in various topographical features, such as the shapes and sizes of the sulcus and gyrus. In other words, this kind of topographical individuality reflects the cytoarchitectonic differences29.
In fact, when classifying the cerebral cortex at the cytoarchitectonic level, each functional region should have a close spatial relationship with the corresponding region defined at the microstructural level30. The definition of the cerebral cortex map based on the equi-territorial hypothesis between functional and microstructural division is not yet sufficiently developed for it to be put to general use. For the time being, the best method of reflecting cytoarchitectonic diversity is to consider the morphological aspects of the gyrus and sulcus. Therefore, we used the ROI based on topographical parcellation, as this provides a viable alternative for the measurement of functional activity. Our method addresses the issue of the brain structure-function relationship in schizophrenia, because the total subregional activity includes certain aspects of the subregional volume. There can be a temporal discrepancy between the structural and functional changes. There have been arguments as to when these structural and functional changes begin in patients with schizophrenia. The 'abnormal neurodevelopment' hypothesis31, which states that problems occur during the prenatal and postnatal neurodevelopment period, has become generally accepted. On the other hand, some researchers have continued to support the 'neurodegenerative' hypothesis32 which states that progressive changes continue after the onset of illness. Our data do not support either of these two hypotheses. Given that the results we obtained for both the mean and total subregional activities were very similar, it is possible to conclude that the prefrontal functional changes in schizophrenia are not affected by the structural volume changes.
It is well known from previous reports that the dorsolateral prefrontal cortex (DLPFC) plays a critical role in working memory8,16. In the current study, however, there was no significant change in the subregional activities within the DLPFC during the working memory condition in either the patient or the healthy group. Stevens et al13. did not find any difference in the DLPFC between the patient and control groups either. Instead, our study showed important differences in the anterior cingulate between the patient and the healthy group. Working memory loading caused increased activity in the right rostral anterior cingulate in the healthy group, but not in the patient group. Moreover, it caused no significant change in the right caudal anterior cingulate in the healthy group, but a decrease in the patient group. Some previous studies have indicated that impaired activation of the anterior cingulate cortex might be the basis for verbal fluency deficit33 or declarative memory34 in schizophrenia. In addition, as shown in Figure 3, the slightly increased activation in the orbitofrontal cortex during working memory in the healthy group was contrasted by a significant decrease in the patient group. Considering that the patient group demonstrated excellent performance without showing any significant difference from the healthy group, this result indicates that patients with schizophrenia have a different activation pattern from that of the control group for working memory. The nature of the functional impairment of working memory in schizophrenia remains obscure, but it can be said that the neural system is abnormally activated when performing the working memory condition. Previous studies have emphasized the role of the DLPFC in this abnormally activated neural system, but the current study suggests that the anterior cingulate and the orbitofrontal cortex also play an important role. It is not yet known whether these differences in functional activation reflect a quantitatively inefficient neural system or an abnormality in terms of quality. It should be mentioned that subtle focal changes might not be detected with the method used in this study. A negative result is possible if activation occurs in a limited area during working memory, instead of in a generalized area of any of the frontal subregions. For example, although focal activity in the supplementary motor area was reduced in the patients with schizophrenia in other studies, in which a different visual working memory task and voxel-based analysis was used10,11, such finding could not be observed in the relatively large ROI that was employed in the current study. Therefore, even if activation did not appear in some subregions in this study, the possibility of regional activation cannot be ruled out. In fact, subtle regional activity changes are well described in another report using voxel-based analysis35.
The primary limitation of this study is the effect of medication. All of the subjects in the patient group were taking various atypical antipsychotics. The confounding effect of antipsychotics medication on memory-related activation of specific brain regions has been suggested in some, although few, previous reports. For example, typical and atypical antipsychotics have different effects on the patient's performance during a cognitive task36. Moreover, it has been reported that risperidone can cause a distortion in the cortical activation changes that occur during working memory tasks in various brain areas, such as the right prefrontal area, the supplementary motor area and the posterior parietal cortex37. As presented in the tables, in the comparison between the two groups in the same task condition, the patient group showed a tendency to have reduced activities in most of the frontal subregions compared to the healthy group. This result may be related to the hypofrontality of schizophrenia, but the possibility of there being a secondary effect caused by the antipsychotic medication cannot be ruled out. However, because the change in functional activation was first observed within the group, the antipsychotic effect was controllable in the between-group difference of the within-group change. Neuroimaging studies of working memory mapping in patients with schizophrenia have proved the existence of 'task-related hypofrontality'10,38. This has the effect of producing artificially poor task performance in certain patients which can be confused with functional impairment39. In fact, poor performance can be due to a lack of effort or motivation, the use of an inappropriate strategy, or difficulty in performing the task40. When the task is too difficult, the subjects utilize cognitive and emotional processes that have no relation to working memory, such as error monitoring, compensatory behavior, withdrawal from the task, the expression of an embarrassed feeling, or proffering a guess. In the current study, however, in order to eliminate these confounding factors as much as possible, we used a relatively easy task and allowed the subjects to have sufficient practice before the actual task performance, and thus the performance of the patient group did not show much difference from the healthy group. In conclusion, to overcome the limitations of previous neuroimaging studies, in the current study, we used exact topographical parcellation to observe the functional abnormality of the prefrontal areas during working memory in schizophrenia. One of the remarkable findings was the reduced functional responses in the right anterior cingulate and in the right orbitofrontal cortex. These abnormalities may be related to the cognitive deficits and impairment of the neural circuits in schizophrenia.
Volz HP, Gaser C, Hager F, Rzanny R, Mentzel HJ, Kreitschmann-Andermahr I, Kaiser WA, and Sauer H: Brain activation during cognitive stimulation with the Wisconsin Card Sorting Test: a functional MRI study on healthy volunteers and schizophrenics. Psychiatry Res: Neuroimaging 1997; 75: 145-157.
Andreasen NC, Rezai K, Allinger R, Swayze VW, Flaum M, Kirchner P, Cohen G, and O'Leary DS: Hypofrontality in neuroleptic-naive patients and in patients with chronic schizophrenia: Assessment with xenon 133 single-photon emission computed tomography and the Tower of London. Arch Gen Psychiatry 1992; 49: 943-958.
Curtis VA, Bullmore ET, Brammer MJ, Wright IC, Williams SCR, Morris RG, Sharma TS, Murray RM, and McGuire PK: Attenuated frontal activation during a verbal fluency task in patients with schizophrenia. Am J Psychiatry. 1993; 155: 1056-1063.
Baddeley A: Working memory. Science 1992; 255: 556-559.
Jonides J, Smith EE, Koeppe RA, Awh E, Minoshima S, and Mintun MA: Spatial working memory in humans as revealed by PET. Nature 1993; 363: 623-625.
D'Esposito M, Detre JA, Alsop DC, Shin RK, Atlas S, and Grossman M: The neural basis of the central executive system of working memory. Nature 1995; 378: 279-281.
Courtney SM, Petit L, Maisog JM, Ungerleide LG, and Haxby JV: An area specialized for spatial working memory in human frontal cortex. Science 1998; 279: 1347-1351.
Smith EE and Jonides J: Storage and executive processes in the frontal lobes. Science 1999; 283: 1657-1661.
Goldman-Rakic PS: Working memory dysfunction in schizophrenia. J Neuropsychiatry Clin Neurosci 1994; 6: 348-357.
Callicott JH, Ramsey NF, Tallent K, Bertolino A, Knable MB, Coppola R, Goldberg T, Van Gelderen P, Mattay VS, Frank JA, Moonen CTW, and Weinberger DR: Functional magnetic resonance imaging brain mapping in psychiatry: methodological issues illustrated in a study of working memory in schizophrenia. Neuropsychopharmacology 1998; 18: 186-196.
Carter CS, Perlstein W, Ganguli R, Brar J, Mintun M, and Cohen JD: Functional hypofrontality and working memory dysfunction in schizophrenia. Am J Psychiatry 1998; 155: 1285-1287.
Menon V, Anagnoson RT, Mathalon DH, Glover GH, and Pfefferbaum A: Functional neuroanatomy of auditory working memory in schizophrenia: relation to positive and negative symptoms. Neuroimage 2001; 13: 433-446.
Stevens AA, Goldman-Rakic PS, Gore JC, Fulbright RK, and Wexler BE: Cortical dysfunction in schizophrenia during auditory word and tone working memory demonstrated by functional magnetic resonance imaging. Arch Gen Psychiatry 1998; 55: 1097-1103.
Manoach DS, Press DZ, Thangaraj V, Searl MM, Goff DC, Halpern E, Saper CB, and Warach S: Schizophrenic subjects activate dorsolateral prefrontal cortex during a working memory task, as measured by fMRI. Biol Psychiatry 1999; 45: 1128-1137.
Elliott R and Dolan RJ: Differential neural responses during performance of matching and nonmatching to sample tasks at two delay intervals. J Neurosci 1999; 19: 5066-5073.
Cohen JD: Temporal dynamics of brain activation during a working memory task. Nature 1997; 386: 604-608.
Kim JJ, Mohamed S, Andreasen NC, O'Leary DS, Watkins GL, Ponto LLB, and Hichwa RD: Regional neural dysfunctions in chronic schizophrenia studied with positron emission tomography. Am J Psychiatry 2000b; 157: 542-548.
Talairach J and Tournoux P: Co-planar stereotaxic atlas of the human brain. Thieme Medical, New York. 1988.
Steinmetz H and Seitz RJ: Functional anatomy of language processing: neuroimaging and the problem of individual variability. Neuropsychologia 1991; 29: 1149-1161.
Kennedy DN, Lange N, Makris N, Bates J, Meyer J, and Caviness Jr VS: Gyri of the human neocortex: an MRI-based analysis of volume and variance. Cereb Cortex 1998; 8: 372-384.
Crespo-Facorro B, Kim JJ, Andreasen NC, O'Leary D, Wiser A, Bailey JM, Harris G, and Magnotta V: Human frontal cortex: an MRI-based parcellation method. NeuroImage 1999; 10: 500-519.
Kim JJ, Crespo-Facorro B, Andreasen NC, O'Leary DS, Zhang B, Harris G, and Magnotta VA: An MRI-based parcellation method for the temporal lobe. NeuroImage 2000a; 11: 271-288.
Crespo-Facorro B, Kim JJ, Andreasen NC, Spinks R, O'Leary DS, Bockholt HJ, Harris G, and Magnotta VA: Cerebral cortex: a topographic segmentation method using magnetic resonance imaging. Psychiatry Res: Neuroimaging 2000; 100: 97-126.
Kim JJ, Kwon JS, Kang DH, Youn T, Han MH, Chang KH: Frontal substructural abnormalities evaluated by the parcellation of MRI in patients with schizophrenia. J Korean Neuropsychiatr Assoc 2002; 41: 998-1019.
First MB, Spitzer RL, Gibbon M, Williams JBW: Structured Clinical Interview for DSM-IV Axis I Disorders. New York: New York State Psychiatric Institute, Biometrics Research. 1996.
Hollingshead AB and Redlich FC: Social Class and Mental Illness: A Community Study. John Wiley and Sons, New York. 1958.
Kay SR: The positive and negative syndromes scale (PANSS) for schizophrenia. Schizophr Bull 1987; 13: 261-267.
Bensaid AM, Hall LO, Bezdek JC, Clarke LP: Partially supervised clustering for image segmentation. Pattern Recognition 1996; 29: 859-871.
Rakic P: Specification of cerebral cortical areas. Science 1988; 241: 170-176.
Roland PE and Zilles K: Structural divisions and functional fields in the human cerebral cortex. Brain Res Rev 1998; 26: 87-105.
Weinberger DR: From neuropathology to neurodevelopment. Lancet 1995; 346: 552-557.
DeLisi LE: Defining the course of brain structural change and plasticity in schizophrenia. Psychiatry Res: Neuroimaging 1999; 92: 1-9.
Dolan RJ, Fletcher P, Frith CD, Friston KJ, Frackowiak RS, and Grasby PM: Dopaminergic modulation of impaired cognitive activation in the anterior cingulate cortex in schizophrenia. Nature 1995; 378: 180-182.
Fletcher P, Buchel C, Josephs O, Friston K, and Dolan R: Learning-related neuronal responses in prefrontal cortex studied with functional neuroimaging. Cereb Cortex 1999; 9: 168-178.
Kim JJ, Kwon JS, Park HJ, Youn T, Kang DH, Kim MS, Lee DS, Lee MC: Functional disconnection between the prefrontal-parietal cortices during working memory processing in schizophrenia: a [15O]H2O PET study. Am J Psychiatry 2003; 160: 919-923.
Braus DF, Ende G, Weber-Fahr W, Sartorius A, Krier A, Hubrich-Ungureanu P, Ruf M, Stuck S, and Henn FA: Antipsychotic drug effects on motor activation measured by functional magnetic resonance imaging in schizophrenic patients. Schizophr Res 1999; 39: 19-29.
Honey GD, Bullmore ET, Soni W, Varatheesan M, Williams SC, and Sharma T: Differences in frontal cortical activation by a working memory task after substitution of risperidone for typical antipsychotic drugs in patients with schizophrenia. Proc Natl Acad Sci USA 1999; 96: 13432-13437.
Yurgelun-Todd DA, Waternaux CM, and Cohen BM: Functional magnetic resonance imaging of schizophrenic patients and comparison subjects during word production. Am J Psychiatry 1996; 153: 200-205.
Ebmeier KP, Lawrie SM, and Blackwood D: Hypofrontality revised: a high resolution single photon emission computed tomography study in schizophrenia. J Neurol Neurosurg Psychiatry 1995; 58: 452-456.
Frith CD, Friston KJ, and Herold S:
Regional brain activity in chronic schizophrenic patients during the performance of a verbal fluency task. Br J Psychiatry 1995; 167: 343-349.