Abstract
Symposium 1 : Early Detection of Psychosis: Role of Electrophysiology
Electrophysiological Markers of Vulnerability to Psychiatric Disorders
Abnormalities of EEG and event-related potentials (ERPs) in neuropsychiatric patients have provided the impetus for studies of potential markers of vulnerability to the development of future psychopathology in a variety of disorders. Sleep EEG patterns have been examined in children of schizophrenic parents. The P300 component of the ERP has been related to future psychopathology in the offspring of schizophrenics, and as a neurodevelopmental risk marker for alcoholism. The P50 component of the evoked potential in sensory gating paradigms has also has been implicated as a risk marker for schizophrenia. This symposium will critically examine the current state of these electrophysiological markers of vulnerability in the offspring of patients, as well as findings for adult relatives.
Event-Related Potential Indices of Cognitive Processing and Familial Risk for Schizophrenia-Spectrum Disorder. Stuart R. Steinhauer, PhD, and Ruth Condray, PhD. Biometrics Research, VA Pittsburgh Healthcare System and Western Psychiatric Institute & Clinic, University of Pittsburgh School of Medicine, USA
Components of the event-related potential associated with salience of information, such as P300 and N400, are reliably observed to be reduced in schizophrenia patients. Moreover, familial similarity has also been established for the morphological characteristics of the ERP waveform. Data to be presented for adult relatives of schizophrenia patients suggest that significant differences for P300 amplitude are not robust in the absence of severe psychopathology. However, language function is impaired in schizotypal individuals who have a family history of schizophrenia, but not in schizotypal individuals without a family history of schizophrenia.
N400 recorded in a language paradigm, in which schizophrenia patients are characterized by abnormalities, will be examined for individuals with schizophrenia-spectrum symptomatology with respect to family history of schizophrenia.
Shirley Y. Hill, PhD.
Electrophysiological markers of vulnerability in child and adolescent offspring of schizophrenia patients. Matcneri Keshavan, MD, Shirley Y. Hill, PhD. Western Psychiatric Institute & Clinic, University of Pittsburgh School of Medicine, USA
Abnormal sleep architecture is seen in many schizophrenia patients, but it is unclear if these represent trait related indices of susceptibility. In ongoing studies, we have obtained evidence suggesting decreased delta sleep in polysomnographic studies in early course schizophrenic patients Delta sleep normally declines during adolescence, and is thought to reflect synaptic pruning processes. It is unclear whether delta sleep alterations may represent trait-related indicators of premorbid vulnerability to schizophrenia; we have conducted sleep EEG studies in a series of child and adolescent offspring (HR-S) and age and sex matched control subjects to address this question. Automated sleep analyses showed significant reductions in average delta counts (.5–3 Hz) among the HR-S subjects. We have also conducted P300 studies in HR-S and age- and sex-matched psychiatrically healthy controls in collaboration with Dr. Shirley Hill.
Visual P300 amplitudes were reduced in the HR-S subjects compared to the control subjects. While preliminary, our findings suggest that delta sleep and visual P300 amplitude reductions are present in at least a subgroup of young individuals at increased genetic risk for schizophrenia.
Lawrence E. Adler, MD, Mental Health Service, Denver VA Medical Center, USA
Electrophysiological Paradigms as Possible Phenotypes for Genetic Linkage:
Lessons Learned from P50 (or, “Lessons Learned the Hard Way!”)
Symposium 2 : Combination of Electrophysiology and Neuroimaging in Psychiatry: Current Status and Perspectives
Oliver Pogarell, D. Rehm, C. Mulert, E. Meisenzahl, S. Teipel, H. Hampel, G. Juckel, U. Hegerl, Munich, Germany: Interhemispheric electroencephalographic coherence and regional corpus callosum area as measures of cortico-cortical connections in Alzheimer's disease and schizophrenia.
Interhemispheric electroencephalographic coherences reflect the integrity of long commissural fibers. We investigated correlations between indices of interhemispheric coherence and regional corpus callosum area in healthy subjects and patients with Alzheimer's disease or schizophrenia as independent functional and structural measures for the integrity of interhemispheric connectivity.
40 patients with schizophrenia, 11 patients with probable Alzheimer's disease (AD), and 40 healthy subjects underwent both magnetic resonance imaging with the assessment of total and regional corpus callosum areas, and quantitative electroencephalography with calculation of regional coherences.
Both patients with AD and schizophrenia showed significant positive correlations between regional interhemispheric coherences and corpus callosum areas, whereas there were no associations between these functional and structural parameters in healthy subjects.
Region specific correlations between corpus callosum size and degree of interhemispheric coherence are suggestive of a reduction in functionally intact commissural fibers under pathological conditions such as AD or schizophrenia, subsequently leading to an impairment in brain connectivity.
Christoph Mulert, P. Bussfeld, L. Jäger, 0. Pogarell, U. Hegerl, G. Juckel, Munich, Germany: New insights into human brain function with simultaneous EEG/fMRI measurements: Focus on target detection using the oddball (P300-) paradigm.
The event related P300 potential, usually evoked with an oddball paradigm, has widely been used in neurophysiological research. The main reason for its broad application is the fact that alterations in the P300 amplitudes and latencies have been described in several brain or mental disorders, such as Alzheimer's disease or schizophrenia. However, a precise correlation of the scalp EEG data by itself to the underlying brain regions is not possible, since the correct localization of the generators of scalp measured data is limited by the low spatial resolution of EEG. With the development of modern imaging technologies like functional Magnetic Resonance Imaging (fMRI), precisely localized brain activations could be detected, but only with restricted temporal resolution. For a more comprehensive understanding of the brain activity underlying the P300 paradigm we therefore have used a combination of EEG and fMRI to achieve both a precise localization and a high time resolution.
Patrick Bussfeld, C. Mulert, H. Brauer, P. Decker, L. Jäger, U. Hegerl, G. Juckel, Munich, Germany: Simultaneous acquisition of 32-channel EEG and fMRI-data during a classical Stroop paradigm.
Simultaneous acquisition of EEG and BOLD sensitive EPI-sequences allows to combine electrical scalp data and BOLD effect derived activation data in order to identify sources of action in the prefrontal cortex during cognitive performance of selective attention.
We measured five healthy female subjects performing an event related Stroop paradigm in a Siemens Vision 1.5T with BOLD sensitive Echo-Planar-Imaging-Sequences, covering the whole head with 10 slices, 8mm slice thickness and a Gap of 0.8mm. TR was 2500ms.
EEG-data were acquired with a Schwarzer 32-channel EEG System (MR-compatible: EMR digital), sampling rate 1000 Hz, high/lowpass-filters 0.53 and 35Hz. EEG acquisition was timed in between scans, leaving 1000ms after each stimulus. The Stroop Paradigm consisted of neutral, incongruent and congruent visual stimuli in a 1:1:1 ratio. Four runs of each 120 stimuli were performed (overall presentation of 480 stimuli).
The fMRI-data were analyzed with BrainVoyager4.4, the EEG-Data with the Brain Vision software. FMRI-data yielded activations in the bilateral mid-prefrontal areas related to incongruent stimuli in contrast to congruent stimuli. Event-related potentials showed different peaks between congruent and incongruent stimuli.
The possible conclusions on the relationsship between fMRI- and EEG-data and the advantages and disadvantages of simultaneous acquisition will be discussed.
Henry H. Holcomb, Baltimore, U.S.A.: Perceptual and Motor Learning in Schizophrenia: Evidence of a Connectional Dysfunction Obtained from Functional Neuroimaging.
Symposium 3 : Sensory Gating: Theoretical, Methodological, and Empirical Advances
Since its introduction to the field two decades ago, the P50 gating paradigm has garnered much interest and significantly added to our understanding of attentional and cognitive deficits in psychiatric populations. This symposium will offer an update of the most recent theoretical, methodological, and empirical findings spanning clinical and electrophysiological levels of analysis. We will present data collected from a variety of psychiatric populations, including schizophrenia, bipolar disorder, anxiety disorders, and autism.
Symposium 4 : Head Injury EEG in Traumatic Brain Injury
The symposium will focus on the clinical sensitivity and specificity of QEEG for the detection of EEG patterns in traumatic brain injured patients. The issues that will be addressed are: 1- The relative sensitivities of EEG power versus EEG coherence and EEG phase in the evaluation of the acute phase and long term follow-up of TBI, 2- The prognostic value of QEEG to predict quality of life one year following a traumatic brain injury, 3- Correlations between EEG and neuropsychological performance in TBI patients, 4- Correlations between EEG and the Glasgow Coma Score in TBI patients and, 5- Cross-validation of the EEG in the diagnosis and evaluation of TBI by correlation with quantitative MRI and 3-D LORETA Neuroimaging.
Symposium 5 : Normative Data-Bases in Neuropsychiatry on the Clinical and Research Utility of a qEEG/EP Database: Neuropsychopharmacological Aspects
As the target organ of a psychotropic drug is the human brain, the EEG seems to be one of the most suitable methods to determine objectively and quantitatively if, how, when, and at what dosage a compound produces an effect on the human CNS. This has been demonstrated initially by visual evaluation, later by computer-assisted quantitative analysis of single-leads, thereafter by multi-lead analysis and mapping techniques and recently by tomographic analyses, which opened the possibility of identifying targets of pharmaceuticals within the brain. Pharmaco-EEG profiles, maps and tomographic images of sedative and non-sedative neuroleptics and antidepressants, anxiolytics and hypnotics, psychostimulants and nootropics have been described. The relationships between acute and chronic drug effects, alterations in normals and patients, central effects and therapeutic efficacy as well as between pharmacokinetics and pharmacodynamics have been discussed. For creating a universal database, however, standardization of data acquisition (e.g., study design, recording condition), data storage (data format for data exchange) and data analysis (e.g., artifact processing, parameterization) is crucial. Many problems have to be considered such as inter- and intraindividual variability factors, age, gender, circadian rhythms as well as normal and pathological states. Only after considering all these factors, a neuropsychopharmacological database will contribute to the clinical and research utilization of EEG/EP.
EEG Normative Databases: Statistical Standards of Sensitivity and Specificity
The statistical standards of Gaussian distributions and cross-validation were applied to EEG samples from 625 normal subjects ranging in age from 2 months to 82 years. Power spectral analyses were computed using Complex Demodulation and the FFT for different EEG frequency bands and different measurement categories of coherence, phase, amplitude asymmetry, total power, absolute power, relative power from 19 scalp locations. Different montages of linked ears, average reference and reference-free current source density or Laplacian norms were independently evaluated. Skewness, kurtosis and Komolgrov-Smirnov estimates of Gaussian distributions were obtained for all variables and logrithmic transforms were used when necessary to more closely approximate a Gaussian distribution. After transforms, all of the variables were shown to be well behaved and approximately Gaussian. An independent cross-validation of each of the 625 subjects with respect to the normative EEG database was conducted by computing Z scores for each of the variables per subject and per age group for linked ears, average reference and the current source density norms. The results of the analyses showed a grand average cross-validation accuracy from 92.3% to 99.2% based upon the approximation to a Gaussian normal distribution. Statistical sensitivity when comparing an individual's EEG values to a normative database is based solely on the extent of the approximation to a Gaussian. Multiple comparison statistics are not possible when comparing an individual to a norm. Plots of sensitivity versus skewness and kurtosis were developed to aid in the statistical evaluation of any normative database measure. Receiver Operating Characteristic curves were also produced to compare the relative accuracyand sensitivity of different EEG measures.
Symposium 6 : Electrophysiologic Studies of Depression: New Findings and Potential Clinical Applications
Studies using quantitative electroencephalography (qEEG) and event-related potential (ERP) measures have found abnormalities in depressed patients and have also found differences between responders and nonresponders to antidepressants. This symposium will focus on recent qEEG and ERP findings, with particular focus on the potential of these measures as predictors of clinical response to treatments for depression.
Loudness Dependence of Auditory Evoked Potentials (LDAEP), Central Serotonergic Activity and Prediction of Clinical Response to Serotonergic Versus Noradrenergic Antidepressants
The central serotonergic system has strong modulatory influences on sensory processing in the primary auditory cortex. The reactivity of the primary auditory processing in the primary auditory cortex can be assessed by studying the N1/P2-component of the auditory evoked potentials with the help of modern source localization programs (Dipole source analysis, LORETA) or with fMRI. Several preclinical and clinical studies have shown, that the loudness dependency of auditory evoked potentials (LDAEP) is one of the best validated indicators of central serotonergic function, 1,2 and that a strong LDAEP indicates a favorable response to serotonin agonists. 3 In a recent study it was not only confirmed that a strong LDAEP predicts a favorable response of depressed patients to SSRI, but also that a low LDAEP is suggestive of a favorable response to the noradrenergic antidepressant reboxetine. The clinical value of these findings will be discussed with special focus on the different techniques of ERP-analysis being used in clinical studies.
Symposium 7 : Event-related Activity in the Human Medial Temporal Lobe: Memory Functions and Malfunctions
Functional integrity of the human hippocampal formation is crucial for the formation and retrieval of long-term memories, and hippocampal dysfunction contributes to neuropsychological deficits in a variety of neurological and psychiatric diseases. Recording event-related potentials (ERPs) from the scalp and directly from within the medial temporal lobes as well as event-related functional magnetic resonance imaging (fMRI) can help to identify the specific hippocampal contributions to memory. This symposium will present both ERP and fMRI data revealing functions of the human medial temporal lobe. Moreover, potential physiological correlates of memory deficits will be discussed by the example of temporal lobe epilepsy.
Declarative Memory Formation Within the Human Medial Temporal Lobe
University of Bonn Bonn, Germany
The integrity of the medial temporal lobe (MTL) is essential for declarative memory. Using functional magnetic resonance imaging we could show that successful encoding into declarative memory engages the posterior hippocampus and the anterior parahippocampal gyrus, which is covered by rhinal cortex. However, the time-course and specific functional significance of these mnemonic operations remained unclear. MTL depth recordings of event-related potentials in epilepsy patients revealed that the rhinal cortex participates early on in semantic processing supporting declarative memory formation indirectly and subsequently the hippocampus in an exclusively mnemonic operation of declarative memory formation. These distinct though interrelated operations are indexed by dissociable differences in event-related potentials arising after about 300 and 500 milliseconds, respectively. Summing-up, declarative memory formation is sequentially organized in a rhinal-hippocampal hierarchy, which is temporarily bound together in one assembly by a transient phase-synchronization in the gamma frequency range of around 40 Hz.
Functional Neuroimaging of the Human Hippocampus: What Have We Learned … and Why Haven't We Learned More?
Functional MRI represents a powerful tool for examining the neural mechanisms of human memory. Previous neuroimaging studies have demonstrated involvement of prefrontal and medial temporal cortical regions (predominantly in the parahippocampal gyrus) in episodic memory, however findings regarding memory-related activity in the hippocampus have been less consistent. I will present findings from event-related fMRI studies examining the pattern of prefrontal and hippocampal activity across episodic memory tasks as well as working memory tasks. A novel finding emerging from these studies is that hippocampal activity reliably correlated with the active maintenance of novel information in working memory. Although the hippocampus is not traditionally associated with working memory, I will argue that all of these findings reflect common aspects of hippocampal coding. Specifically, these findings suggest that neural computations in the hippocampus (through interaction with prefrontal cortex) may support the formation, maintenance, manipulation, and retrieval of new memories.
Brain Activity Evidence That the Hippocampal Formation is Critical for Recollection but Not for Familiarity
Several cognitive models of recognition memory converge on the notion that normal recognition has two qualitatively different bases, recollection and familiarity. On the basis of multivariate analyses (partial least squares) of electromagnetic time-average and frequency data, evidence will be provided in favor of this distinction in healthy subjects. It will be shown that the electromagnetic index of recollection-based recognition is absent in a patient with relatively isolated bilateral hippocampal damage whose parahippocampal region appears to be bilaterally intact. Furthermore, ERP data obtained in healthy aged adults suggest that the degree of hippocampal atrophy (volumetry) and cell loss (1 H-spectroscopy) during aging is correlated with a decrease in the ERP index of recollection-based but not familiarity-based recognition. Together, these data indicate that the hippocampal formation is critical for recollection but not for familiarity.
Limbic Event-Related Potentials and Memory Deficits in Temporal Lobe Epilepsy
Mesial temporal lobe epilepsy (MTLE) is associated with typical memory deficits. Depth recordings of limbic event-related potentials (ERPs) can help to shed light on psychophysiological mechanisms of these deficits. Limbic ERPs can be recorded when the presurgical evaluation of patients with temporal lobe epilepsy necessitates the implantation of depth electrodes. By this means we found that both entorhinal and hippocampal ERPs tapping specific memory processes are altered by the epileptogenic process indicating that both encoding and recollection of verbal memories are impaired in MTLE. Moreover, our data suggest that hippocampal NMDA-receptor dependent long-term potentiation (LTP) may be requisite for the detection of relative or situational “novelty” which, in turn, is necessary for the formation of verbal memory entries. In clinical practice limbic ERPs can be used to predict postoperative verbal memory performance in MTLE patients after hippocampectomy in the dominant temporal lobe.
Symposium 8 : Gelastic Seizures
This symposium will bring together basic and clinical experts in the field of epilepsy, genetics, developmental biology, neuroradiology and neurosurgery for an intensive effort for the characterization of this rare epileptic syndrome, to deveop standards of managements and treatments and to identify future research directions.
Symposium 9 : Funding Opportunities
Symposium 10 : ERP Analyses, Advanced Methodology Advanced Methods for EP Analysis
Recent advances in recording technology, along with new techniques for biosignal analysis, provide an unprecedented opportunity to study human brain dynamics. This symposium will present emerging computational tools that can be used to analyze and interpret EEG and EP activity obtained during performance of a variety of experimental tasks involving sensory perception and higher cognitive function, such as memeory, language, and attention.
Non-Stationary Signal Analysis Tools to Study Evoked Potential Generation†
Traditionally, the spontaneous EEG has been viewed as “noise”, obscuring the much smaller evoked potential. However, there is mounting evidence that the EP reflects a reorganization of this spontaneous activity, rather than an independent, additive process. This reorganization of spontaneous EEG may result in increased activity in certain frequency bands, or phase reorganization. In fact, the possibility that the EP reflects a modification of the relative phases of the spectral components present in the spontaneous activity was raised almost thirty years ago, but the lack of adequate signal analysis tools prevented early researchers from drawing firm conclusions regarding the mechanisms underlying EP generation.
In this paper, we will present advanced signal analysis tools developed for single trial EP analysis. When these tools were applied to auditory EPs, we uncovered evidence that mid-latency auditory evoked potential components are generated by a different mechanism than the late components, such as P300 obtained with an auditory odd-ball experiment.
Footnotes
†
This research was supported in part by grants R01 MH58784, and THECB-003652-0001-1999