W. Grey Walter

Comparison of subcortical, cortical and scalp activity using chronically indwelling electrodes in man

Abstract 1. 1. Intracerebral recordings show that there is little field spread (volume conduction) in brain tissue and that cerebral models with dipoles deep within the brain are not satisfactory. 2. 2. Comparison of subdural and scalp recordings shows that only widely synchronised components of the cortical activity are observed on the scalp. 3. 3. For strictly localised activity the attenuation from cortex to scalp can be as high as 5000 : 1, but for coherent activity over a wide area it may be only 2 : 1. 4. 4. Model experiments indicate that cortical areas of at least 6 sq cm must be involved in synchronous or near synchronous activity before the scalp EEG is observed, using standard working gains. 5. 5. A plea is made for higher working gains for part of routine EEG recordings.

The central effects of rhythmic sensory stimulation

Abstract 1. 1. Records obtained during photic stimulation may be described in terms of the following components, any or all of which may be present at any one time, or from time to time. 1.1. A. A series of dicrete elementary evoked responses, e.g. figures 1B and 8. 1.2. B. Fusion of evoked responses giving an accidental appearance of rhythimicity (first part of figure 5). 1.3. C. Instrumental summation of evoked response and spontaneous rhythms (Fig. 1A). 1.4. D. True augmentation of “driving” of local rhythms at the frequency of the stimulus. 1.5. E. Augmentation of harmonically related rhythms in other areas. 2. 2. Differences between individuals are attributable in some cases to anatomical variations and correlate also to some extent with the character of their spontaneous activity, with age and with differences in personality. 3. 3. Alterations in the response in given individuals are produced by somatic, mental and emotional changes whether spontaneous, voluntary or induced. 4. 4. Somatic, mental and emotional changes can be induced in the subject by stimulation at appropriate frequencies. 5. 5. The above effects can interact with one another in both regenerative and degenerative fashion. 6. 6. Subjective visual effects are attributed to interference between rhythmic evoked responses, and spontaneous rhythms at cortical and possibly thalamic levels. 7. 7. Anomalous (non-visual) effects in normal and abnormal between rhythmic evoked responses to interaction between the evoked activity and harmonically related spontaneous rhythms in other circuits at a thalamic level. 8. 8. Evocation of activity in non-visual circuits can be used to study their physiology and as an to diagnosis of some pathological conditions. 9. 9. Some theoritical implications of these findings are speculatively discussed.

Regional control of cerebral vascular reactivity and oxygen supply in man

Abstract Gold electrodes implanted for treatment of psychiatric disorders and intractable epilepsy were used to measure the local oxygen availability (O2a) in the intact brain. Relative measurement of local blood flow was obtained from micro-thermistors implanted with the electrodes. Recordings were taken during hyperpnoea, hypercapnia, hypoxia and breathing 100% oxygen. Changes of O2a and blood flow were observed during visual and electrical stimulation, changes of posture and heart irregularities. Results are compatible with a regional auto-regulation of local cerebral oxygenation. It is suggested that the Bohr shift of the oxyhaemoglobin dissociation curve by metabolic and arterial CO2 determines the cerebral tissue oxygenation during short-lived changes of metabolic rate and respiratory variations.

A new toposcopic display system.

Abstract 1. 1. The advantages and disadvantages of voltage-time recording are discussed. 2. 2. A method is described whereby changes of voltage from 24 channels are displayed in a spatial co-ordinate system. 3. 3. Methods are described for indicating frequency, phase and time relations in each channel. 4. 4. The applications of the method are briefly described with reference to model generators and the study of evoked activity.

Intracerebral events in humans related to real and imaginary stimuli

Abstract Recordings were made from gold intracerebral electrodes in orbital and cingulate cortex and subdural electrodes on superior frontal cortex in five patients under treatment for chronic obsessional and anxiety disorders. The subjects were instructed to guess whether or not they would receive a stimulus in the form of a flash, click or electric shock to the finger. They were told to move a lever to the right when they expected a stimulus and to the left when they did not. A reward was offered for every three successive correct guesses. Movement of the lever in either direction started a LINC 8 averaging and provided a stimulus or not according to a switching programme operated by the experimenter. Sets of averages of eight trials each were collected in the four conditions: expectation-occurrence, expectation-non-occurrence, no expectation-occurrence, no expectation-non-occurrence. Averages were also taken of responses to random stimuli and of intrinsic activity when the lever was moved without guessing or stimuli. Clear cerebral events, termed “emitted potentials”, were observed when stimuli were expected but did not occur. The emitted potentials resembled those evoked when real stimuli were presented on switch closure, suggesting that they may reflect memory processes corresponding to the perception of real events. In some experiments the latency of the emitted potentials was significantly shorter than that of responses evoked by real random stimuli, suggesting a rise in cortical excitability with expectancy. A positive deflection peaking at about 300 msec often appeared following switch closure, particularly with expectancy, and a negative variation usually preceded and accompanied this action.

Contingent negative variation and evoked responses recorded by radio-telemetry in free-ranging subjects

Abstract 1. 1. Using an 8-channel radio-telemetry system (RTE) records were obtained of the EEG, pulse rate, respiration, evoked responses and Contingent Negative Variation (CNV) in four normal subjects and three patients with intracerebral electrodes. The subjects were free to move about within 30 m of the receiving aerial. Two RTE channels were modified to provide time constants of 7 sec. 2. 2. Auditory stimuli, synchronized with the operation of average response computers, were transmitted to the subjects by a separate radio-control link. These were used as conditional and imperative signals to the subjects to perform various tasks: pressing a button to arrest the imperative signals, turning the pages of a book, inserting pegs in a peg-board and coming to a stop on a bicycle. 3. 3. Responses to the signals were averaged on line with two barrier-grid tubes and a 2-channel Enhancetron and the intrinsic rhythms were analysed with a 2-channel frequency analyser. The pulse rate was indicated by a cardiotachometer. 4. 4. The initil responses and CNV during the reception of paired auditory signals were similar to those seen with direct connection, provided that the subjects were engaged in some task related to the signals. The CNV was attenuated during exercise or conversation only while the subject was inattentive to the signals. 5. 5. The amplitude of the intracerebral responses to the conditional and imperative stimuli in the patients with implanted electrodes was reduced when the scalp CNV was attenuated by isolation or distraction. 6. 6. During the performance of fairly complex tasks following the auditory signals, the CNV terminated only at the completion of the task, not at the moment of muscular effort. 7. 7. The radio-control link was also used to instruct an experimenter when to loss a ball to a subject or to feint. In this situation also the CNV developed only when the subject was sure the ball was in the air and terminated when it was caught. 8. 8. These observations suggest that the interactions of evoked responses and CNV seen in laboratory conditions also accompany normal activity and the performance of everyday tasks.

Emitted cerebral events

Abstract EEG scalp records from human subjects were used to determine if (a) there were emitted cerebral events, similar to evoked potentials which occured when stimuli were expected but absent, and (b) of there were slow waves similar to the contingent negative variation (CNV) which preceded and followed feedback information about response performance. A paradigm was used which included 3 conditions, all of which were modifications of the standard CNNV procedure. The subjects were required to press a button within a time-window which was automatically varied according to their reaction time. In some conditions S2 was omitted randomly and in one condition there was no manual response to the imperative stimulus, but an EEG change which occured within the window was defined as a response. In all conditions subjects were given feedback information about response performance 1.5 sec after the imperative stimulus. The results were analysed using a new pattern recognition technique which did not require averaging of raw data. The results indicated that there were emitted events at the approximate time when the imperative stimulus should have occured on those occasions when it was absent. The emitted events correlated well with potentials evoked by the imperative stimulus when it was present. the results also showed that there were slow waves preceding feedback and following feedback which correlated with the preceeding S1–S2 CNV. These results were discussed in relation to the nature of registration and retrieval mechanisms and also in relation to the findings reported by others of persistent negatively after the CNV in subjects with behaviour pathology.

A technique for the measurement of phase relations of the EEG

Abstract A method is described in which the sine and cosine components of the Fourier analyses of multichannel data are displayed as vectors. Common reference recordings are recommended because of the interaction of amplitude and phase differences in bipolar montages.

Discussion sur les bases mathématiques de l'analyse de fréquence

Abstract In EEG we are dealing with a variable quantity which is a message containing information. To analyze its frequencies is to break it down into its components in order to make it more useful than in its original form. The method of analysis must (1) apply to all possible messages, (2) loose nothing of the information, (3) make the most useful transformation. Mackays theorem states that no information is lost in a reversible transformation. The method of analysis should make up a “biunivoque” transformation of the value containing the message. The only way to accomplish this is to break the message down into a sequence of orthogonal functions which form a complete unit in the functional space considered. Now the brain, functioning with expenditure of energy, may be considered as a dynamic system, i.e. functioning in a state of equilibrium with its surrounding. It thus seems appropriate to use as sequences of orthogonal functions the trigonometric functions—i.e. sine and cosine values of the small oscillations about the point of equilibrium— and to analyze the message by Fourriers series.