Electroencephalography and clinical Neurophysiology, 1984, 58: 7P-10P
Elsevier Scientific Publishers Ireland, Ltd. Society proceedings EASTERN
New York, NY, December 1, 1983
Dr. IRA SHERWIN
200 Springs Road (151), Bedford, MA 01730 (U.S.A.) (Received for publication: January 27, 1984)
1. Discovery of hidden data in the EEG by power spectral density and pattern recognition algorithm. - - R. Cohn (Howard University College of Medicine, Washington, DC) The respiratory and cardiac rhythms are putatively constant, generally hidden, components of the EEG in man and animals. Work was done in the gerbil to determine the conditions under which the respiratory and cardiac rhythms could unequivocally be expressed in the power spectral density (PSD) and a pattern recognition algorithm (ALG) in synoptic displays. Using the transduced respiratory and cardiac rhythms to synchronize and to summate the time-locked EEG activity, these rhythms were proved to be constant components of all EEGs. Despite the incontestable presence of the respiratory and cardiac rhythms they were not recognized as distinct components in either the PSD or ALG. Under conditions of cerebral ischemia the respiratory and cardiac activities are transiently the prime electrical phenomena in the EEG; at such times the relative recognition merits of PSD and ALG were again compared. Now both the PSD and ALG showed definite recognition of the respiratory and cardiac activities. From the experimental data it is concluded that neither the PSD nor the ALG generate incisive recognition of EEG constituents that are not observable in the on-going EEG. 2. Muscimol in substantia nigra: proconvulsant to anticonvulsant effect with maturation. - - S.L. Moshe and B.J. Albala (Albert Einstein College of Medicine, Bronx, NY) Since recent data utilizing GABAergic stimulation of the substantia nigra (SN) suggest that the SN is a crucial site in a circuitry involved in the suppression of seizures in adult rats, the role of SN in seizures in rat pups was investigated. In adult rats bilateral nigral infusions of the GABA-agonist muscimol increased the latency to the onset of flurothyl induced convulsions (g + SE = 14.9 + 1.1 rain) 1 h after the completion of the infusions as compared to controls (8.1+0.5 min, P < 0.01). This protective effect of muscimol was not evident 24 h later with the latency to the onset of the convulsions decreasing to control levels (muscimol rats = 9.3 + 0.6 min, controls 9.3 + 0.3 min). Similar infusions of muscimol actually facilitated the development of flurothyl seizures in 15-day-old rat pups with a
mean latency to the onset of seizures 7 + 0.4 min as compared to 8.8+0.7 rain in controls ( P < 0.01). Furthermore 24 h later the mean latency of the muscimol infused pups significantly increased to 10+0.7 min ( P < 0.01) and now did not differ from the controls (10.2+1.4 min). These results suggest that the increased seizure susceptibility of the immature brain may be due to developmental differences in the nigral GABA-sensitive system. 3. Piezo crystals in sleep polygraphy. Recordings of uncontaminated electro-oculograms. - - L. H611er (Berlin, F.R.G.) Piezo crystals transform mechanical pressure into voltage differences and also possess pyro-electrical (heat-sensitive) properties. Hence piezo crystals may be used for the recordings of (a) eye movements, (b) body movements, (c) swallowing movements, (d) finger pulses, (e) tremor and (f) respiration (thoracic, abdominal and nasal air flow). We used a piezo element for polygraphic nocturnal sleep studies (from Siemens-Elema, No. 230) weighing 0.5 g and thus easily attachable to the eye lid where it maintains its correct location for the duration of sleep without any discomfort. In comparison with conventional canthus-glabella electrodes, the piezo crystal records ocular movements without any EEG contamination. It also shows some oculographic detail not found in the conventional EOG. The piezo crystal shows special versatility in the recording of respiratory functions. It documents movements of chest, abdomen and nostrils but also records rapid temperature fluctuations of the nasal air flow. This technical innovation has proved to be very helpful in nocturnal sleep research and is likely to be valuable in the routine EEG laboratory. 4. Differential effect of anticonvulsant drugs on excitatory and inhibitory mechanisms. - - G.H. Fromm, C.F. Terronce and A.S. Chatta (University of Pittsburgh, Pittsburgh, PA) Anticonvulsants probably act by preventing the spread and generalization of the paroxysmal discharge from the epileptogenic focus to surrounding normal neuronal systems. We have therefore compared the effect of carbamazepine (CBZ), valproate (VPA) and ethosuximide (ESM) on excitatory and
0013-4649/84/$03.00 © 1984 Elsevier Scientific Publishers Ireland, Ltd.
8P inhibitory mechanisms in the trigeminal complex of cats. CBZ depressed afferent and periventricular descending excitatory pathways, and periventricular inhibition, but facilitated segmental inhibition. VPA depressed periventricular excitatory and inhibitory pathways, but had little effect on afferent excitation or segmental inhibition. ESM depressed periventricular and segmental inhibition, but had no effect on afferent or periventricular excitation. Our results indicate that drugs effective against absence seizures (VPA, ESM) depress reticular inhibitory pathways. This suggests that such seizures are due to paroxysmal discharges in inhibitory pathways. Drugs effective against generalized tonic clonic seizures (CBZ, VPA) depress reticular excitatory pathways and thus prevent the generalization of paroxysmal activity via the reticular core. Drugs that are also effective against partial seizures (CBZ) additionally facilitate negative feedback systems, thereby limiting the spread of paroxysmal activity from the epileptogenic focus to surrounding neuronal systems.
SOCIETY PROCEEDINGS patients, to determine the role of AEEG in initial assessment of episodes which might be seizures. Seven patients had poorly characterized spells and normal neurologic examinations, 6 with normal EEGs and 1 with diffuse slowing; all AEEGs were unrevealing. Six had episodes consistent with complex partial seizures, of whom 2 had diffuse slowing and 2 focal abnormality; 1 patient had a seizure in AEEG and 3 had episodes without AEEG concomitants. Nine patients had previously documented seizures of different type, 6 with abnormal examinations, 6 with focal EEG abnormality, and 2 with diffuse slowing; 4 had seizures on AEEG, 3 interictal discharges, 1 an episode without concomitants and 1 an unrevealing recording. Three patients had normal examinations, psychiatric symptoms, and described atypical consistent with pseudoseizure, 2 with normal EEGs and 1 with diffuse slowing: all three had multiple clinical episodes without AEEG concomitants. The findings suggest that: AEEG is unhelpful as a screening test, can detect pseudoseizures, and can confirm interictal discharges in conventional EEG proven cases of epilepsy.
5. Myoclonic seizures in childhood: an analysis using EEG FM radiotelemetry and videotape recording. - - M. McKeever and G.L. Holmes (University of Connecticut Health Center, Farmington, CT)
7. Falling asleep: depth EEG and thermographic observations. - - E. Niedermeyer and W.R. Jankel (Johns Hopkins Hospital, Baltimore, MD)
In order to accurately evaluate the clinical and EEG manifestations of myoclonic seizures in children, 182 seizures recorded in 16 pediatric patients (mean age = 9.25 years) during EEG FM radiotelemetry and videotape monitoring were reviewed. Myoclonic seizures were diagnosed using criteria outlined by Gastaut (1970). Infantile spasms and myoclonic jerks which occurred as components of other seizures were excluded. All 16 patients had abnormal interictal EEGs with the majority demonstrating multifocal spikes, polyspikes, or polyspike and wave activity. Clinically myoclonic seizures were unilateral in 82/182 (45.1%), bilateral in 53/182 (29.1%), generalized in 25/182 (13.7%) and involved the head only in 22/182 (12.1%). The electroencephalographic correlate consisted of generalized spike or polyspike and wave activity in 152/182 (83.5%) and generalized polyspike activity in 30/182 (15.5%). There was a poor correlation between EEG abnormalities and the clinical manifestations of the seizures. Four patients had a marked deterioration of their EEG with an increase in spikes, polyspikes and polyspike and wave activity for several minutes following the myoclonic seizure. Overall the myoclonic seizures were very stereotyped in both clinical and electroencephalographic manifestations.
The appearance of sleep spindles indicates the onset of light NREM sleep (stage 2). This rule, however, does not apply to the depth EEG. Thalamic implants (14 patients with intractable chronic pain) showed independent ventrobasal thalamic spindle activity which, in 3 patients, occurred while the scalp record showed patterns of early drowsiness. In frontal and limbic implants in 78 patients with intractable epileptic seizures (mostly temporal lobe epileptics), spindles were noted in superior frontal regions (supplementary motor zone) while most of the patients showed the scalp EEG patterns of light drowsiness or even wakefulness with posterior basic rhythm. These observations suggest that the process of falling asleep starts in wakefulness or very light drowsiness as determined by behavior and scalp EEG. Portions of the frontal lobe and the thalamus apparently initiate this process. There is also an additional peripheral component which expresses itself as the well known sensation of 'heavy eye lids,' burning and itching around the eyes. At this stage, thermography shows ocular and periocular hot spots which resolve in sleep. The slow process of falling asleep is likely to depend on neural and humoral mechanisms.
6. Ambulatory cassette EEG monitoring in initial assessment of suspected seizures. - - M.E. Drake, Jr. (Ohio State University, Columbus, OH)
8. Electrocorticographic considerations of the epileptogenic focus in man. - - E. Niedermeyer, M. Graf, S. Uematsu and D.M. Long (Johns Hopkins Hospital, Baltimore, MD)
Ambulatory cassette EEG monitoring (AEEG) has been found useful in quantifying seizures of interictal discharges in known epileptic patients, and as an alternative to prolonged inpatient EEG monitoring. We investigated the relationship between initial clinical evaluation and AEEG findings in 25
This study is based on 100 intra-operative electrocorticographic recordings obtained in 96 patients with chronic severe epileptic seizure disorder. Most of these patients underwent temporal lobectomy. Special emphasis was placed on the spatial characteristics of
EASTERN EEG SOCIETY, 37th A N N U A L M E E T I N G spike foci. During preceding scalp EEG recordings, focal spike or sharp wave activity was noted in most cases. Recordings from the cortex, however, showed, in the majority of cases, the presence of several foci of spike activity, synchronously or independently firing, sometimes rather widely scattered over the explored cortical area. Independent limbic spiking was sometimes demonstrable with additional one-lead-depthelectrodes. In a minority of the cases, the circumscript character of the spike focus was demonstrable on the scalp as well as on the exposed cortex. The type of anesthesia (general, local) did not significantly alter these findings. These data suggest a dissipation of 'focal' epileptogenic activity into several sub-foci in the majority of temporal lobe epileptics. There is reason to presume that, in most cases of temporal lobe epilepsy, large portions of the lobe are involved (neocortical and limbic). Because of the 'focus dissipation,' lobectomies might be preferable to small excisions.
9. Could interictal epileptic spikes actually be post-ictal epileptic spikes? - - J. Gotman (Montreal Neurological Institute, McGill University, Montreal, Canada) Seizures and interietal spikes, because they both originate in the same epileptogenic region, are usually believed to share the same pathophysiological mechanisms. The frequency of appearance of both can be extremely variable but the reasons for these fluctuations are largely unknown (except for sleep which increases interictal spiking). Using a computer system for long-term monitoring and quantification of interictal spikes and sharp waves, we studied the relationships between interictal activity and the occurrence of seizures in patients having scalp or intracerebral electrodes, and in kindled cats. EEGs were analyzed during several days and sometimes weeks. Results indicated that: (i) no specific change in interictal activity was seen prior to seizure occurrence; (ii) some seizures were followed by a decrease in spiking lasting a few hours; (iii) most seizures were followed by a large increase in spiking starting either immediately or a few hours after the seizure and lasting from one to several days. It therefore appears that seizure occurrence is a major cause of increased spiking. Could it be that seizures are actually the cause of interictal spiking? Spikes could then reflect the damage caused to the brain by seizures and would more appropriately be labelled post-ictal.
10. Neuronal activity during ictal-interictal cycles in primary and secondary epileptogenic loci. - - I. Sherwin (E.N.R. Memorial Veterans Hospital, Bedford, MA) Previously we reported that major differences exist in the fine-structure (intraburst) firing patterns recorded interictally, from penicillin primary and secondary (mirror) epileptogenic focus neurons. These are present even when, by EEG criteria, the foci may appear to be independent of one another. In the current study an attempt was made to assess the influence of ictal discharge in modifying the behavior of secondary focus neurons compared to interictal discharge alone.
9P Each of the cells studied was classified according to its behavioral correlation with the field potentials, i.e., surface paroxysms (SPs): type I, increased activity; type II, decreased activity; type III, no clear correlation. Secondary focus type I and type lI neuronal behavior was primarily characterized by suppression of activity. Such cells also showed stronger correlations with SPs occurring in the primary focus rather than the secondary focus itself. The behavior of type Ill cells was analyzed utilizing difference-spectra. This revealed that primary focus type Ill units are characterized by greater internal coherence of their inter-paroxysms spike trains. The results of the study indicate that for acute epileptogenic foci repeated seizures, as compared to interictal discharges alone, do not constitute a more potent factor for modifying the behavior of secondary focus neurons. These findings have relevance for kindling and chronic secondary epileptogenesis.
1 i. Stability in waking-sleep states in neonates as a predictor long-term neurological outcome: further development. - C.T. Lombroso and Y. Matsumiya (Children's Hospital Medical Center, Boston, MA) Thirty full-term newborns were studied. These newborns were ranked by 3 pediatricians from 'normal' to 'at risk' for future neurological/developmental outcomes. The time spent in each of 3 stages of waking and 2 stages of sleep were recorded in each newborn for 4 consecutive weeks. Thus individual profiles of each baby's ability to maintain fairly constant state organization were obtained. The 'consistency profile' was compared with the long-term outcome of each child. It was found that the individual state consistency over 4 weeks would provide a high degree of predictability on the future neurological outcome. Neither the initial clinical evaluation by pediatricians, nor the result of the organization of a single waking or sleep state could produce a comparably high degree of predictability.
12. Digital filtering of BAEPs and SSEPs recorded from scalp and spinal cord. - - P.J. Maecabee, N. Hassan, R.Q. Cracco and J.A. Schiff (SUNY Downstate Medical Center, Brooklyn, NY) BAEPs (6 subjects) and median nerve SSEPs (20 subjects) were performed utilizing open bandpass and a non-cephalic reference. Intraoperative spinous process-reference recordings were performed in 5 patients at open bandpass. Recordings were then digitally filtered utilizing selected high or low pass filter settings. BAEPs at low pass (5-250 c/sec) revealed a diffuse scalp positivity (onset 2 - 3 msec, peak 5 - 6 msec) upon which were superimposed faster frequency Jewett waves when recorded at open bandpass. SSEPs at low pass (1-100 c/sec) revealed a diffuse scalp positivity (onset 7 msec, peak 11-12 msec) which was obscured by P l l and P13 when recorded at open bandpass. Peripheral nerve action potentials recorded below 100-150 c / s e c were severely attenuated. Fast potentials recorded from L 3, T12 and T 6 spinous processes were preserved at bandpasses
10P above 300 c/sec. The slow large negative potential at TI2 was fractionated into 2 negative peaks with high pass filtering and the following slow positive potential was severely attenuated. We postulate that evoked potential components may be properly 'tuned' according to their fundamental spectral composition. Depending upon the specific system under study, selected bandpass may possibly distinguish components arising predominantly within axonal a n d / o r fast synaptic generators from those arising in slower synaptic, asynchronous axonal, or glial generators. 13. Evoked potential testing in acute coma in childhood: etiologic and prognostic considerations. - - Y. Matsumiya, V. Gibbons, R. Sawa and C.T. Lombroso (Children's Hospital Medical Center, Boston, MA) Sixty children and young adults in coma of recent onset were tested. The age of the patients ranged from 9 months to 24 years. Most patients had only brain-stem auditory evoked potentials (BAEP) but some also had brain-stem and cortical somatosensory a n d / o r cortical visual evoked potentials. It was found that if the first BAEP was severely abnormal, all patients died. In cases in which at least waves I, 1II and V were visible, it was found that both anoxic and metabolic groups have longer wave I I I - V conduction time than the other groups such as infections and trauma. In metabolic and anoxic groups, if the first BAEP was found normal, the mortality was low. Serial recordings are found to be very important. If BAEPs in serial recordings worsened, prognosis was extremely poor. Conversely, their improvement was associated with fair prognosis. A new prognostic method using I - I I I / I I I - V conduction ration was developed. This method seems to have some advantages over the classical methods in which only raw amplitude a n d / o r latency values are used. In some cases in which coma is caused by bihemispheric lesions, cortical evoked potentials of other modalities were demonstrated to be useful. 14. Scalp topography of human somatosensory evoked potentials; effects of interfering tactile stimulation applied to the hand. - - C.N. Power and S. Jones (Univ. of Ottawa Med. School, Ottawa, Canada) A topographical analysis of human SEPs in response to median nerve stimulation was performed', with and without concurrent tactile stimulation of the hand. The study employed 28 recording sites on the neck and scalp with non-cephalic as well as earlobe reference electrodes. The following components
SOCIETY PROCEEDINGS were observed in the control response: N10, N l l , NI3, PI4, N17, N19, P20, P22, N22, N29, P30, and P44. The interference response showed significant amplitude diminution of P14, N 19, P20, P22 and N29, while N10, N l l , N13 and N22 remained unchanged, P30 was abolished and P44 was enhanced. In the interference response there also appeared an N33 component, supplanting P30 over the contralateral post-central region. A 'difference' wave form, obtained by subtracting the interference from the control response, included ' P'I4, 'N'19, ' P'21, ' P'32 and 'N'36 components. The last 2 were of large amplitude over the contralateral post-central region and frontal cortex respectively, and overlapped for at least 15 msec of their duration. The results indicate that the interfering effect occurred at the brain-stem level or higher. The difference wave form suggests that the 'P'32 and 'N'36 components (partially masked) arise from a generator in Brodmann's area 3b, and the P22 and N22 components of the control and interference wave forms from Brodmann's area 3a. 15. Rapid visual assessment from the evoked potential: an alternative to computer averaging. - - M. Kupersmith, J. Nelson, W. Seiple and R. Carr (NYU Medical Center, New York, NY) We have extended and refined an electrophysiological technique which allows a real-time lock-in retrieval of the pattern visual evoked potential (PVEP). This allows key visual performance parameters to be evaluated with a continuously changing visual display. The classical averaged PVEP is an objective measure of neural pathway conduction in the visual system but has given limited data which can be related to visual threshold performance or visual complaints. Averaging methods are too slow to allow a complete assessment of the visual spatial-temporal domain in patients. We have successfully applied a rapid technique to obtain an electrophysiologically defined contrast sensitivity function by continuously increasing (sweeping) contrast from sub- to suprathreshold levels over a 20 sec epoch for each spatial frequency of interest. In another application, sweeping the spatial frequency from course to fine gratings, with contrast held constant, has allowed us to measure an objective equivalent of Snellen acuity. In clinical application, we have identified hysterical amblyopia, uncovered selective contrast losses in multiple sclerosis, and quantified acuity loss in macula degeneration and optic neuropathy. Thus, our refined, rapid method permits an objective functional dissection of the visual system which was previously only available using highly subjective psychophysical techniques.