Responses to Rare Visual Target and Distractor Stimuli Using Event ...

RAPID COMMUNICATION

Responses to Rare Visual Target and Distractor Stimuli Using Event-Related fMRI VINCENT P. CLARK,1,4 SEAN FANNON,4 SONG LAI,2,4 RANDALL BENSON,3,4 AND LANCE BAUER1 Departments of 1Psychiatry, 2Diagnostic Imaging and Therapeutics, 3Neurology, and 4Program in Functional NeuroImaging, University of Connecticut Health Center, Farmington, Connecticut 06030-2017 Clark, Vincent P., Sean Fannon, Song Lai, Randall Benson, and Lance Bauer. Responses to rare visual target and distractor stimuli using event-related fMRI. J. Neurophysiol. 83: 3133–3139, 2000. Previous studies have found that the P300 or P3 event-related potential (ERP) component is useful in the diagnosis and treatment of many disorders that influence CNS function. However, the anatomic locations of brain regions involved in this response are not precisely known. In the present event-related functional magnetic resonance imaging (fMRI) study, methods of stimulus presentation, data acquisition, and data analysis were optimized for the detection of brain activity in response to stimuli presented in the three-stimulus oddball task. This paradigm involves the interleaved, pseudorandom presentation of single block-letter target and distractor stimuli that previously were found to generate the P3b and P3a ERP subcomponents, respectively, and frequent standard stimuli. Target stimuli evoked fMRI signal increases in multiple brain regions including the thalamus, the bilateral cerebellum, and the occipital-temporal cortex as well as bilateral superior, medial, inferior frontal, inferior parietal, superior temporal, precentral, postcentral, cingulate, insular, left middle temporal, and right middle frontal gyri. Distractor stimuli evoked an fMRI signal change bilaterally in inferior anterior cingulate, medial frontal, inferior frontal, and right superior frontal gyri, with additional activity in bilateral inferior parietal lobules, lateral cerebellar hemispheres and vermis, and left fusiform, middle occipital, and superior temporal gyri. Significant variation in the amplitude and polarity of distractor-evoked activity was observed across stimulus repetitions. No overlap was observed between target- and distractor-evoked activity. These event-related fMRI results shed light on the anatomy of responses to target and distractor stimuli that have proven useful in many ERP studies of healthy and clinically impaired populations.

INTRODUCTION

The process of visual target detection in the presence of interleaved distractor stimuli generates event-related potential (ERP) components that have been found to be clinically useful in the diagnosis and treatment of Alzheimer’s disease, schizophrenia, depression, closed head injury, and drug dependence, among many other CNS disorders (Bauer 1997; Polich 1999). Rare distractor stimuli that evoke the frontal P300 or P3a subcomponent have proven especially useful in the study of disorders thought to involve prefrontal and limbic cortex dysfunction (Bauer 1997; Knight and Nakada 1998). Although the exact cause of the P3 component latency and amplitude differences found in these clinical populations is not fully understood, it is thought to involve deficits in arousal, attention, The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

stimulus processing speed, and/or the memory operations required to perform these tasks (Polich and Kok 1995). Functional magnetic resonance imaging (fMRI) could provide an improved measure of the specific neural fields that may be involved in this response by virtue of the greater anatomic specificity of fMRI when compared with ERPs. This in turn would assist in the interpretation of these ERP effects in clinical populations and may offer additional information useful for the diagnosis and treatment of CNS disorders. Previous ERP studies suggested that the oddball task involves multiple neural fields that support the cognitive processes necessary for its performance. These include occipital, temporal, parietal, frontal, thalamic, cerebellar, and limbic regions that respond to rare target stimuli, and dorsolateral prefrontal, medial prefrontal, parietal, and posterior hippocampal regions that respond to rare distractor stimuli (Baudena et al. 1995; Courchesne et al. 1995; Halgren et al. 1998; Polich and Kok 1995; Swick and Knight 1998). In contrast, previous fMRI studies using similar stimuli found target-evoked activity in a less extensive range of brain regions (McCarthy et al. 1997; Menon et al. 1997; Opitz et al. 1999) and few regions that were identified for rare distractor stimuli (Kirino et al. 1997; Knight and Nakada 1998). This suggests that some portion of the neural activity evoked by these stimuli is not observed using fMRI. In a previous study (Clark et al. 1998), we introduced a method for performing event-related fMRI using multiple regression, which has shown greater sensitivity to task-related signal change than other methods. In an effort to identify additional target- and distractor-evoked activity using eventrelated fMRI, this method was adapted here to identify brain regions that respond to a three-stimulus visual oddball task based on Bauer (1997). Importantly, we hypothesized repetition-related changes in response amplitude such as might occur with habituation or changes in response strategy, and which were previously reported for the P300 ERP response (Knight 1984; Knight and Scabini 1998, Ravden and Polich 1998). Using these methods, target- and distractor-evoked activity was found in a wide range of brain regions. METHODS

Stimuli were frequent standard (the letter “T,” P ⫽ 0.82), rare distractor (the letter “C,” P ⫽ 0.09), and rare target (the letter “X,” P ⫽ 0.09) single block-letter stimuli presented sequentially in pseudorandom order for 200 ms each with the interstimulus interval (ISI) varied randomly from 550 to 2050 ms across trials. Subjects were instructed to make a speeded button-press response upon each pre-

0022-3077/00 $5.00 Copyright © 2000 The American Physiological Society

3133

3134

V. P. CLARK, S. FANNON, S. LAI, R. BENSON, AND L. BAUER

FMRI

TABLE

3135

STUDY OF THREE-STIMULUS ODDBALL TASK

1. Areas with significant main effect of stimulus presentation

ROI Center (Brodmann Area) Standard stimulus Right superior temporal gyrus (38) Cerebellum Anterior cingulum (24) Left middle occipital gyrus (18) Target stimulus Thalamus Medial lingual gyrus (18) Left fusiform gyrus (19, 37) Right fusiform gyrus (19, 37) Left superior temporal gyrus (22) Left postcentral gyrus (1, 2, 3, 40) Right inferior parietal lobule (7, 40) Right middle frontal gyrus (6, 9) Medial frontal gyrus (32, 6) Left insula (13) Left inferior frontal gyrus (9) Right precentral gyrus (44)

ROI Range (Brodmann Area)

Cuneus (17, 18) Cerebellum, lingual gyrus (18), middle occipital gyrus (18, 19, 37) Cerebellum, middle occipital gyrus (18, 19, 37) Inferior parietal lobule (40), middle temporal gyrus (21), supramarginal gyrus (40) Inferior parietal lobule (40) Superior temporal gyrus (22), postcentral gyrus (1, 2, 3, 5), supramarginal gyrus (40) Inferior frontal gyrus (9), precentral gyrus (4, 6, 9) Cingulate gyrus (24, 32), superior frontal gyrus (6, 8) Precentral gyrus (13, 44, 6), inferior frontal gyrus (45, 47) Inferior frontal gyrus (45, 47)

Volume (cc)

Coordinates (x,y,z)

1.12 1.42 1.03 1.54

34,20,⫺28 ⫺18,⫺85,⫺44 ⫺3,33,5 ⫺36,⫺88,0

2.26 2.30 13.2

⫺1,⫺12,11 ⫺1,⫺76,2 ⫺38,⫺60,⫺16

5.68 3.60 15.5 13.4 6.34 14.2

29,⫺52,⫺14 ⫺61,⫺45,21 ⫺47,⫺24,53 48,⫺39,38 32,0,56 ⫺3,10,46

4.71

⫺45,11,5

1.44 6.47

⫺64,19,15 44,17,6

ROI, region of interest.

sentation of the target letter. Subjects were further requested to maintain gaze. Direction of gaze was not monitored. The distractor letter had the same presentation frequency as the target letter, but no response was required. Each run was comprised of 90 stimuli. A minimum of eight runs was acquired per subject. Stimulus sequence development is described in detail in Clark et al. (1998). Briefly, a large number of random stimulus sequences were generated. Sequences with a low correlation of predicted blood oxygen level– dependent (BOLD) responses among the three stimulus types (¦r¦ ⬍ 0.2) were used. This allowed the BOLD response evoked by each stimulus type to be tested separately using multiple regression. Six healthy right-handed volunteers (2 females, 4 males; mean age 32 yr, range 21– 44 yr) participated in this study. A gradient echo, echo planar scanning sequence was used [single-shot, 20 oblique slices 6-mm thick, no skip, repetition time (TR) ⫽ 2150 ms, echo time (TE) ⫽ 40 ms, flip angle ⫽ 90°, field of view ⫽ 25.6 cm, matrix ⫽ 64 ⫻ 64] on a 1.5 Tesla Vision magnetic resonance imaging (MRI) scanner (Siemens, Erlangen, Germany). High-resolution three-dimensional inversion recovery MPRAGE volumes were also acquired in the same orientation and field of view. Movement correction and spatial normalization were performed using SPM96b (Friston et al. 1995). A modified version of the

Talairach coordinate system (Talairach and Tournoux 1988) was used. The significance of stimulus-evoked changes in MRI signal intensity was evaluated using multiple regression (Haxby et al., in press). This analysis was identical to that used in Clark et al. (1998) except that regions of interest (ROIs) from group-averaged data were located by identifying clusters of 10 or more contiguous voxels significant at P ⬍ 0.001 (Friston et al. 1994). Additional analyses to detect the presence of linear trends in response amplitude over stimulus repetitions were performed using two additional regressors to test for linear trends. The analyses were performed at a higher voxel-wise significance threshold (P ⬍ 0.0001) to compensate for the use of additional comparisons. To confirm the results of these statistical tests, stimulus onset time-locked averages (analogous to the ERP) were obtained from detrended fMRI data and averaged across voxels within individual ROIs. RESULTS

The mean accuracy for detection of the target stimulus was 95% (reaction time ⫽ 466 ⫾ 110 ms, SD). Less than one false alarm was made per subject on average. Standard and target stimuli evoked a significantly increased BOLD signal in a total

FIG. 1. Areas of significant activation of spatially normalized data illustrate the anatomic regions of activity. All images are displayed according to radiological convention. Numbers indicate positions of slices in the normalized z-axis dimension. The group-averaged statistical map is superimposed over the high-resolution structural data of 1 representative subject. A: responses to standard stimuli. ¦Z¦ ⬎ 3.09; region of interest (ROI), P ⬍ 0.005. B: response to target stimuli plotted as in A. C: regions with significant trend in distractor response amplitude across successive distractor repetitions. ¦Z¦ ⬎ 4.0; ROI, P ⬍ 0.005. D: medial frontal gyrus ROI for successive distractor repetitions. ¦Z¦ ⬎ 4.0; ROI, P ⬍ 0.005. Plotted in sagittal, coronal, and axial orientations. E: superior frontal gyrus ROI for successive target repetitions plotted as in D. FIG. 2. A: stimulus time locked averages shown for 5 ROIs from the group-averaged data that responded to the target stimulus. From top to bottom, the plots were taken from ROIs centered in the thalamus (THAL), left occipitotemporal cortex/cerebellum (LOTC), left inferior parietal gyrus (LIP), right middle frontal gyrus (RMF), and medial frontal/cingulate gyri (MFC). Amplitude of percentage change relative to 6.45 s prestimulus baseline is shown. B: target stimulus time locked averages obtained from each of the 6 subjects (S1–S6) plotted in red to purple for the same five ROIs shown in A. C: plots depicting changes in group mean blood oxygen level– dependent signal intensity as a function of repeated stimulus presentations for the medial frontal gyrus ROI shown in Fig. 1D. Values represent the peak percentage change of signal intensity from baseline per stimulus presentation. Responses to distractor and target stimuli are shown in red and blue, respectively. D: plotted as in C for the ROI illustrated in Fig. 1E.

3136


of 5 and 89 ml of brain tissue, respectively (Fig. 1, A and B; Table 1). Target stimuli generated activity near the regions identified in previous fMRI studies (McCarthy et al. 1997; Menon et al. 1997; Opitz et al. 1999), including the thalamus, bilateral inferior parietal lobules, cingulate, and right middle frontal gyri. Additional activity was observed bilaterally in the cerebellum; the middle occipital, lingual, and fusiform gyri; the medial occipital cortex; the bilateral precentral, postcentral, insular, medial frontal, polar superior frontal, and inferior frontal gyri; and the left middle temporal gyrus. The larger volume of target-evoked activity observed in the present study compared with these previous studies may be caused by the faster stimulus presentation rate, the larger imaging volume, and the use of multiple regression as employed here; the significance of target-evoked activity was not reduced by the use of fewer subjects. No regions of significantly decreased BOLD signal were observed, which is in agreement with previous fMRI studies using the oddball task (McCarthy et al. 1997). No significant response was observed for the main effect of distractor presentation. Instead, a significant positive trend in signal intensity across distractor repetitions was found in a total of 7 ml of tissue, with the largest volume of significant response located in the prefrontal cortex (Fig. 1, C and D; Table 2). A smaller volume (0.84 ml) was observed for target repetitions in the medial prefrontal cortex (Fig. 1E). 2. Areas with significant trend in response amplitude across successive stimulus repetitions

TABLE

ROI Center (Brodmann Area)

ROI Range (Brodmann Area)

Volume (cc)

Coordinates (x,y,z)

Positive trend Distractor stimulus Medial frontal gyrus (10) Anterior cingulum (32) Left inferior frontal gyrus Right inferior frontal gyrus (47) Right polar superior frontal gyrus (10) Left cerebellum Right cerebellum Cerebellar vermis Left anterior fusiform gyrus (20) Left middle occipital gyrus (18) Left superior temporal gyrus (22) Left inferior parietal lobule (40, 7) Right inferior parietal lobule (40, 7) Target stimulus Superior frontal gyrus (32) Medial frontal gyrus (11, 12)

ROI, region of interest.

DISCUSSION

With careful attention to techniques of stimulus presentation, data acquisition, and data analysis, BOLD fMRI responses to target and distractor stimuli can be observed in brain regions that overlap those reported previously using similar statistical thresholds (Kirino et al. 1997; Knight and Nakada 1998; McCarthy et al. 1997; Opitz et al. 1999) and in additional regions thought to be involved in the P3 ERP response for similar stimuli (Baudena et al. 1995; Courchesne et al. 1995; Halgren et al. 1998; Polich and Kok 1995; Swick and Knight 1998). Importantly, the use of multiple regression analysis in the present study to examine higher-order BOLD fMRI responses led to the finding of significant distractor-evoked activity in widespread brain regions, including ventral-medial prefrontal regions previously observed using depth-recorded ERPs (Baudena et al. 1995) but not using fMRI. Comparison with P300 ERP studies

3.73

⫺8,47,0

0.14

⫺34,33,⫺16

0.27

27,17,⫺22

0.82 0.58 0.28 0.34

9,70,10 ⫺62,40,⫺46 29,⫺71,⫺53 0,⫺60,⫺46

0.15

⫺38,⫺13,⫺26

0.06

⫺47,⫺77,⫺8

0.10

⫺69,⫺39,16

0.15

⫺27,⫺63,31

0.26

24,⫺50,20

0.70

⫺16,52,⫺13

0.14

5,58,⫺16

Negative trend Distractor stimulus Right fusiform gyrus (37)

Averaged BOLD time series showed that response amplitudes differed between stimulus types in a manner consistent with the statistical analyses (Fig. 2A). The consistency of the target response was examined across subjects (Fig. 2B), which revealed activity for all subjects in all regions except the left inferior frontal gyrus ROI, where two of the six subjects did not show a response. A reversal from a negative response evoked by the first stimulus in each run to a positive response evoked near the end of the run was observed for ROIs with significant positive trends (Fig. 2, C and D). The significant linear trend in response amplitude across repetitions was found to be stimulus specific within each ROI, showing that this effect is not caused by nonspecific changes in response amplitude or baseline shift.

0.09

47,⫺33,⫺10

The presentation of rare target and nontarget distractor stimuli evokes a class of ERP components, collectively termed the P300 or P3 component, that have been well studied since their discovery (Sutton et al. 1965). Indeed, over 1,500 manuscripts included in the Medline database since that time make specific reference to this component in either their titles or abstracts. The importance of this phenomenon is further illustrated by its application in a wide range of clinical ERP studies, including most CNS disorders (Polich 1999). The P300 can be divided into at least two major subcomponents, the P3a and the P3b. The P3b is evoked by the presentation of target stimuli, to which subjects make a motor response or perform a mental instruction such as counting. This component is largest near the temporal-parietal junction and is thought to index the attention- and memory-related operations involved in stimulus processing. The infrequent presentation of distractor stimuli that interrupt the performance of the primary task produces an earlier positive potential (the P3a) that is largest over frontal and central scalp regions and that habituates rapidly (Comerchero and Polich 1998).1 This is believed 1 Various types of distractor stimuli were used in previous P300 ERP studies. These may be differentiated depending on whether the distractor stimuli are “novel” and differ categorically from the standard and target stimuli or, instead, if they fall into the same stimulus category as the other stimuli, as in the present study. Distractor stimuli may also be differentiated based on whether individual instances of specific distractor stimuli are presented only once or whether the same distractor stimulus is presented multiple times

FMRI


to reflect an alerting or orienting response that may originate in frontal cortex. Finding distractor-evoked activity in a large portion of the prefrontal cortex remained elusive in previous fMRI studies but agrees with previous ERP findings of medial and inferior lateral prefrontal activity that is evoked by rare distractor stimuli (Baudena et al. 1995) and eliminated by lesions of the prefrontal cortex (Knight 1984). The present results support the hypothesis that distractor stimuli evoke activity in prefrontal cortical regions and further indicate that additional brain regions are involved, including the medial and lateral cerebellum, bilateral inferior parietal lobules, and the left occipitaltemporal cortex. This suggests the possibility that reductions in P3a amplitude observed in patients with CNS disease might be caused by the dysfunction of brain regions outside the prefrontal and limbic cortices. A significant linear trend was observed in the amplitude of distractor-evoked activity, which reversed from negative to positive on average (relative to the prestimulus baseline) during the course of each experimental run. This finding may be caused by habituation of the neural response across stimulus repetitions, as was described for the amygdala (Breiter et al. 1996; Buchel et al. 1998). However, this is in contrast to typical ERP findings where the P3a component evoked by similar distractor stimuli begins with a large positive amplitude that diminishes toward zero over repeated stimulus presentations (Comerchero and Polich 1998). This discrepancy may be caused by differences between the mechanisms of ERP and BOLD fMRI signal generation, which have yet to be fully described, or by neural activity that might possibly be recorded using fMRI but is not observed in the typical ERP, such as event-related desynchronization (Pfurtscheller and da Silva 1999). Such differences could lead to the sampling of activity generated by different neural fields, thus resulting in a different pattern of response. This finding could also reflect an alteration in response strategy with repeated stimulus presentations, resulting in a concomitant change in metabolic demand. In general, the present results point to the importance of examining higher-order response characteristics in neurophysiological data (Friston et al. 1998). Relationship between cognitive processes and associated neural activity Although the three-stimulus oddball task used here is relatively simple in design, many cognitive operations are required for its accurate performance, including response inhibition, preparation and execution, vigilance (or sustained attention), object recognition, working memory, and long-term memory (Polich and Kok 1995). Self-monitoring of response accuracy and changes in mood and arousal (including the orienting response) would also be expected to occur. Because most perceptual and cognitive processes are neurally mediated by multiple brain regions rather than by a single area (Posner and Raichle 1993), the performance of this task in healthy popu(Knight 1984). Differences in the topography and latency of the P3 subcomponent evoked by these various types of distractor stimuli (Polich 1999) using different sensory modalities (Swick and Knight 1998) have been observed. However, most distractor stimuli have been found to generate a P3 subcomponent that peaks earlier and is situated more anteriorly than the target-evoked P3b and that is generally referred to as the P3a subcomponent.

3137

lations would be expected to generate activity within a diverse set of brain regions that support these many cognitive operations. Although the exact correspondence between the cognitive operations required to perform this task and the specific brain regions involved is beyond the scope of the present study, it would generally be expected that response preparation and execution would involve the somato-motor cortex and that vigilance, response inhibition, and working memory would involve the frontal and parietal cortices. Activity was observed within these regions as expected. Additionally, the analysis of letter identity would be expected to occur in the occipitaltemporal cortex for all stimulus types. The only region found to have overlapping activity for the main effect of standard and target stimulus presentation, as well as an adjacent linear trend for distractor stimuli, was observed in the left middle occipital gyrus (see slices located at z ⫽ 0 mm in Fig. 1, A, B, and D), which is near regions observed in other studies using block letter stimuli (Puce et al. 1996). However, no individual voxels showed overlap among the three stimulus types at the statistical thresholds used here, suggesting that once the target identification stage is completed, the neural and cognitive processes invoked in response to this information diverge significantly. This also indicates that neither stimulus frequency nor stimulus response requirements alone can account for these results as both distractor and target stimuli were presented with equal frequency and both the distractor and standard stimuli did not require a response. Instead, it is the combination of stimulus frequency and task demands that determines which brain regions respond to these stimuli. Interpretation of the present results can also be assisted by making comparisons with the results of previous brain imaging studies for anatomic similarity. For instance, distractor-evoked activity was observed in medial prefrontal, inferior parietal, and cerebellar regions in the present study; altered activity in these regions was also observed in depressed patients (Bench et al. 1992; Dolan et al. 1992; Mayberg et al. 1999) and in patients experiencing induced sadness (George et al. 1995; Mayberg et al. 1999). This suggests that the neural fields that respond to distractor stimuli may be associated with variations in mood related to the three-stimulus oddball task or with cognitive processes that are shared by both changes in mood and the detection of rare distractor stimuli. These medial and inferior-lateral prefrontal regions are also near orbitofrontal areas found to be involved in decision making (Bechara et al. 1998). The anterior brain regions (anterior cingulate and insulae) that responded to the target stimulus in the present study overlap with regions consistently found in studies of pain perception and aversive conditioning (Buchel et al. 1998; Rainville et al. 1997). The anterior cingulate is involved in attention, motivation, and behavioral regulation (Devinsky et al. 1995). This region has reciprocal connections with the anterior insula, which is involved in processing emotionally relevant contexts (Fink et al. 1996), among other functions. This network has been hypothesized to provide a mechanism through which nociceptive inputs are integrated with memory to allow appropriate responses to stimuli that predict future adversities (Buchel et al. 1998). The present results suggest that this combination of activity may not be uniquely specialized for the processing of aversive stimuli, but that it may

3138


include the processing of behaviorally relevant nonaversive stimuli as well. Clinical applications of fMRI studies using the threestimulus oddball task The task used here was based on that used by Bauer (1997) for patients recovering from cocaine dependency. Bauer’s study found that the P3b component evoked by the target stimulus was significantly smaller shortly after the cessation of cocaine use and that it then returned to normal after two months of abstinence. One possible explanation for this finding is that there is an overlap between brain regions found to respond to the target stimulus in the present study and those that were found to bind cocaine (Telang et al. 1999). These regions include the thalamus, cingulate gyri, insulae, calcarine cortex, precentral gyrus, lateral frontal cortex, and cerebellum. These regions could experience the residual effects of cocaine use that reduce P3b amplitude, only to normalize during sustained abstinence. Bauer also showed that the amplitude of the P3a component evoked by the rare distractor stimulus identified 71% of the cocaine-dependent individuals who later relapsed to cocaine use. This P3a amplitude reduction predicted relapse with a higher degree of accuracy than any other variable tested in the study. The present findings suggest that this effect may involve the medial and inferior prefrontal cortices and/or the posterior regions found in the present study. Because a variety of cognitive processes are required for the oddball task to be accurately performed, it has proven to be useful in assessing neural and cognitive function in patients with CNS disease. The present results show that in healthy subjects a variety of brain regions respond to target and distractor stimuli in this task. Obtaining similar fMRI data from clinical populations could be useful for the diagnosis of CNS disease, especially in its early stages. Cognitive deficits resulting from CNS illnesses might be associated with different neurophysiological responses in the brain regions that support these cognitive processes. Obtaining fMRI data may also prove useful in identifying patient subgroups, such as was accomplished previously using P300 ERPs in cocaine dependence (Bauer 1997) and schizophrenia (Turetsky et al. 1998). Using fMRI would allow such subgroups to be identified with greater anatomic precision. Such information would be useful for evaluating the effectiveness of different treatment methods among patient subgroups identified in this manner. We are very grateful to J. Lackey, T. Wojtusik, Drs. S. Luck, J. Polich, J. Townsend, D. Swick, and J. Nuetzel, and to the staff of the Dept. of Diagnostic Imaging and Therapeutics at the University of Connecticut Health Center for their assistance. Address for reprint requests: V. P. Clark, Dept. of Psychiatry, MC2017, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06032. Received 4 August 1999; accepted in final form 11 January 2000.

REFERENCES BAUDENA, P., HALGREN, E., HEIT, G., AND CLARKE, J. M. Intracerebral potentials to rare target and distractor auditory and visual stimuli. III. Frontal cortex. Electroenceph. Clin. Neurophysiol. 94: 251–264, 1995.

BAUER, L. O. Frontal P300 decrements, childhood conduct disorder, family history, and the prediction of relapse among abstinent cocaine abusers. Drug Alcohol Depend. 44: 1–10, 1997. BECHARA, A., DAMASIO, H., TRANEL, D., AND ANDERSON, S. W. Dissociation of working memory from decision making within the human prefrontal cortex. J Neurosci. 18: 428 – 437, 1998. BENCH, C. J., FRISTON, K. J., BROWN, R. G., SCOTT, L. C., FRACKOWIAK, R. S., AND DOLAN, R. J. The anatomy of melancholia–focal abnormalities of cerebral blood flow in major depression. Psychol. Med. 22: 607– 615, 1992. BREITER, H. C., ETCOFF, N. L., WHALEN, P. J., KENNEDY, W. A., RAUCH, S. L., BUCKNER, R. L., STRAUSS, M. M., HYMAN, S. E., AND ROSEN, B. R. Response and habituation of the human amygdala during visual processing of facial expression. Neuron 17: 875– 887, 1996. BUCHEL, C., MORRIS, J., DOLAN, R. J., AND FRISTON, K. J. Brain systems mediating aversive conditioning: an event-related fMRI study. Neuron 20: 947–957, 1998. CLARK, V. P., MAISOG, J. M., AND HAXBY, J. V. An fMRI study of face perception and memory using random stimulus sequences. J. Neurophysiol. 79: 3257–3265, 1998. COMERCHERO, M. D. AND POLICH, J. P3a, perceptual distinctiveness, and stimulus modality. Cognit. Brain Res. 7: 41– 48, 1998. COURCHESNE, E., AKSHOOMOFF, N. A., TOWNSEND, J., AND SAITOH, O. A model system for the study of attention and the cerebellum. In: Perspectives of Event-Related Potentials Research, edited by G. Karmos, M. Molnar, V. Csepe, I. Czigler, and J. Desmedt. Amsterdam: Elsevier, 1995. DEVINSKY, O., MORRELL, M. J., AND VOGT, B. A. Contributions of anterior cingulate cortex to behavior. Brain 118: 279 –306, 1995. DOLAN, R. J., BENCH, C. J., BROWN, R. G., SCOTT, L. C., FRISTON, K. J., AND FRACKOWIAK, R. S. Regional cerebral blood flow abnormalities in depressed patients with cognitive impairment. J. Neurol. Neurosurg. Psychiatry 55: 768 –773, 1992. FINK, G. R., MARKOWITSCH, H. J., REINKEMEIER, M., BRUCKBAUER, T., KESSLER, J., AND HEISS, W. D. Cerebral representation of ones own past neural networks involved in autobiographical memory. J. Neurosci. 16: 4275– 4282, 1996. FRISTON, K. J., ASHBURNER, J., FRITH, C. D., POLINE, J-B., HEATHER, J. D., AND FRACKOWIAK, R.S.J. Spatial registration and normalization of images. Hum. Brain Map. 3: 165–189, 1995. FRISTON, K. J., FLETCHER, P., JOSEPHS, O., HOLMES, A., RUGG, M. D., AND TURNER, R. Event-related fMRI: characterizing differential responses. NeuroImage 7: 30 – 40, 1998. FRISTON, K. J., WORSLEY, K., FRACKOWIAK, R.S.J., MAZZIOTTA, J. C., AND EVANS, A. C. Assessing the significance of focal activations using their spatial extent. Hum. Brain Map. 1: 210 –220, 1994. GEORGE, M. S., KETTER, T. A., PAREKH, P. I., HORWITZ, B., HERSCOVITCH, P., AND POST, R. M. Brain activity during transient sadness and happiness in healthy women. Am. J. Psychiatry 152: 341–351, 1995. HALGREN, E., MARINKOVIC, K., AND CHAUVEL, P. Generators of the late cognitive potentials in auditory and visual oddball tasks. Electroenceph. Clin. Neurophysiol. 106: 156 –164, 1998. HAXBY, J. V., MAISOG, J. M., AND COURTNEY, S. M. Multiple regression of effects of interest in fMRI time series. In: Mapping and Modeling the Human Brain, edited by P. Fox, J. Lancaster, and K. Friston. New York: Wiley. In press. KIRINO, E., BELGER, A., GORE, J. C., GOLDMAN-RAKIC, P. S., AND MCCARTHY, G. A comparison of prefrontal activation to infrequent visual targets and non-target novel stimuli. Soc. Neurosci. Abs. 23: 493, 1997. KNIGHT, R. T. Decreased response to novel stimuli after prefrontal lesions in man. Electroenceph. Clin. Neurophysiol. 59: 9 –20, 1984. KNIGHT, R. T. AND NAKADA, T. Cortico-limbic circuits and novelty. Rev. Neurosci. 9: 57–70, 1998. KNIGHT, R. T. AND SCABINI, D. Anatomic bases of event-related potentials and their relationship to novelty detection in humans. J. Clin. Neurophysiol. 15: 3–13, 1998. MAYBERG, H. S., LIOTTI, M., BRANNAN, S. K., MCGINNIS, S., MAHURIN, R. K., JERABEK, P. A., SILVA, J. A., TEKELL, J. L., MARTIN, C. C., LANCASTER, J. L., AND FOX, P. T. Reciprocal limbic-cortical function and negative mood: converging PET findings in depression and normal sadness. Am. J. Psychiatry 156: 675– 682, 1999. MCCARTHY, G., LUBY, M., GORE, J., AND GOLDMAN-RAKIC, P. Infrequent events transiently activate human prefrontal and parietal cortex as measured by functional MRI. J. Neurophysiol. 77: 1630 –1634, 1997.

FMRI


MENON, V., FORD, J. M., LIM, K. O., GLOVER, G. H., AND PFEFFERBAUM, A. Combined event-related fMRI and EEG evidence for temporal-parietal cortex activation during target detection. Neuroreport 8: 3029 –3037, 1997. OPITZ, B., MECKLINGER, A., VON CRAMON, D. Y., AND KRUGGEL, F. Combining electrophysiological and hemodynamic measures of the auditory oddball. Psychophysiology 36: 142–147, 1999. PFURTSCHELLER, G. AND LOPES DA SILVA, F. H. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin. Neurophysiol. 110: 1842–1857, 1999. POLICH, J. P300 in clinical applications. In: Electroencephalography, Basic Principles, Clinical Applications, and Related Fields, edited by E. Niedermeyer and F. Lopes da Silva. Baltimore: Urban and Schwarzenberg, 1999, p. 1073–1091. POLICH, J. AND KOK, A. Cognitive and biological determinants of P300. Biol. Psychol. 41: 103–146, 1995. POSNER, M. AND RAICHLE, M. Images of Mind. New York: McGraw Hill, 1993. PUCE, A., ALLISON, T., ASGARI, M., GORE, J. C., AND MCCARTHY, G. Differential sensitivity of human visual cortex to faces, letterstrings, and textures: a functional magnetic resonance imaging study. J. Neurosci. 16: 5205–5215, 1996.

3139

RAINVILLE, P., DUNCAN, G. H., PRICE, D. D., CARRIER, B., AND BUSHNELL, M. C. Pain affect encoded in human anterior cingulate but not somatosensory cortex. Science 277: 968 –971, 1997. RAVDEN, D. AND POLICH, J. Habituation of P300 from visual stimuli. Int. J. Psychophysiol. 30: 359 –365, 1998. SUTTON, S., BRAREN, M., ZUBIN, J., AND JOHN, E. R. Evoked potential correlates of stimulus uncertainty. Science 150: 1187–1188, 1965. SWICK, D. AND KNIGHT, R. Cortical lesions and attention. In: The Attentive Brain, edited by R. Parasuraman. Cambridge: MIT Press, 1998, p. 143–162. TALAIRACH, J. AND TOURNOUX, P. Co-Planar Stereotaxic Atlas of the Human Brain. New York: Thieme Medical, 1988. TELANG, F. W., VOLKOW, N. D., LEVY, A., LOGAN, J., FOWLER, J. S., FELDER, C., WONG, C., AND WANG, G. J. Distribution of tracer levels of cocaine in the human brain as assessed with averaged [11C] cocaine images. Synapse 31: 290 –296, 1999. TURETSKY, B. I., COLBATH, E. A., AND GUR, R. E. P300 subcomponent abnormalities in schizophrenia. I. Physiological evidence for gender and subtype specific differences in regional pathology. Biol. Psychiatry 43: 84 –96, 1998.