Moshe Abeles

Physiological aspects of information processing in the basal ganglia of normal and parkinsonian primates

There are two views as to the character of basal-ganglia processing - processing by segregated parallel circuits or by information sharing. To distinguish between these views, we studied the simultaneous activity of neurons in the output stage of the basal ganglia with cross-correlation techniques. The firing of neurons in the globus pallidus of normal monkeys is almost always uncorrelated. However, after dopamine depletion and induction of parkinsonism by treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), oscillatory activity appeared and the firing of many neurons became correlated. We conclude that the normal dopaminergic system supports segregation of the functional subcircuits of the basal ganglia, and that a breakdown of this independent processing is a hallmark of Parkinsons disease.

Multispike train analysis

Multispike trains are encountered often, either purposely or inadvertently, when one records from neural populations. This paper focuses on techniques for detecting and identifying the spikes in multispike trains. Relatively simple methods are briefly reviewed. Most of these require a high signal-to-noise ratio. A method based on signal detection by template matching, which works well with relativeiy small spikes, is described in detail. Use of this technique is illustrated by an investigation of the biophysical aspects of extracellular recording in sensory cortex. A further application is the analysis of multi-unit records to display relationships between two or three neurons recorded simultaneously.

Quantification, smoothing, and confidence limits for single-units' histograms.

In this article the relationships among firing rate, probability of firing and counts per bin are examined. It is suggested that PSTHs, autocorrelations and crosscorrelations of neuronal activity should all be expressed in units of firing rates (spikes/s), since the values obtained by such scaling are independent of bin size and of total time of measurement. A simple method for these histograms is described. Methods to compute confidence limits for PSTHs, autocorrelations and crosscorrelations are suggested. The computations are based on the null hypothesis that the spike train(s) is (are) the realization of (independent) Poisson-point process(es). The validity and the limitations of these computations methods, when applied to spike trains, are discussed. Methods to smooth out random fluctuation with little distortion of the histograms shape are described. It is suggested that one can minimize the distortion of the histogram in the time-domain and in the frequency-domain by using a bell-shaped bin whose center point slides continuously along the histogram. The article aims at giving the potential user of the methods some insight for the meaning of the formulae. It describes in detail how the methods are applied in practice and illustrates each method by using real data from single-unit recordings.

Predicting Movement from Multiunit Activity

Previous studies have shown that intracortical activity can be used to operate prosthetic devices such as an artificial limb. Previously used neuronal signals were either the activity of tens to hundreds of spiking neurons, which are difficult to record for long periods of time, or local field potentials, which are highly correlated with each other. Here, we show that by estimating multiunit activity (MUA), the superimposed activity of many neurons around a microelectrode, and using a small number of electrodes, an accurate prediction of the upcoming movement is obtained. Compared with single-unit spikes, single MUA recordings are obtained more easily and the recordings are more stable over time. Compared with local field potentials, pairs of MUA recordings are considerably less redundant. Compared with any other intracortical signal, single MUA recordings are more informative. MUA is informative even in the absence of spikes. By combining information from multielectrode recordings from the motor cortices of monkeys that performed either discrete prehension or continuous tracing movements, we demonstrate that predictions based on multichannel MUA are superior to those based on either spikes or local field potentials. These results demonstrate that considerable information is retained in the superimposed activity of multiple neurons, and therefore suggest that neurons within the same locality process similar information. They also illustrate that complex movements can be predicted using relatively simple signal processing without the detection of spikes and, thus, hold the potential to greatly expedite the development of motor-cortical prosthetic devices.

On the Transmission of Rate Code in Long Feedforward Networks with Excitatory–Inhibitory Balance

The capability of feedforward networks composed of multiple layers of integrate-and-fire neurons to transmit rate code was examined. Synaptic connections were made only from one layer to the next, and excitation was balanced by inhibition. When time is discrete and the synaptic potentials rise instantaneously, we show that, for random uncorrelated input to layer one, the mean rate of activity in deep layers is essentially independent of input firing rate. This implies that the input rate cannot be transmitted reliably in such feedforward networks because neurons in a given layer tend to synchronize partially with each other because of shared inputs. As a result of this synchronization, the average firing rate in deep layers will either decay to zero or reach a stable fixed point, depending on model parameters. When time is treated continuously and the synaptic potentials rise instantaneously, these effects develop slowly, and rate transmission over a limited number of layers is possible. However, the correlations among neurons at the same layer hamper reliable assessment of firing rate by averaging over 100 msec (or less). When the synaptic potentials develop gradually, as is the realistic case, transmission of rate code fails. In a network in which inhibition only balances the mean excitation but is not timed precisely with it, neurons in each layer fire together, and this volley successively propagates from layer to layer. We conclude that the transmission of rate code in feedforward networks is highly unlikely.

Modeling compositionality by dynamic binding of synfire chains.

This paper examines the feasibility of manifesting compositionality by a system of synfire chains. Compositionality is the ability to construct mental representations, hierarchically, in terms of parts and their relations. We show that synfire chains may synchronize their waves when a few orderly cross links are available. We propose that synchronization among synfire chains can be used for binding component into a whole. Such synchronization is shown both for detailed simulations, and by numerical analysis of the propagation of a wave along a synfire chain. We show that global inhibition may prevent spurious synchronization among synfire chains. We further show that selecting which synfire chains may synchronize to which others may be improved by including inhibitory neurons in the synfire pools. Finally we show that in a hierarchical system of synfire chains, a part-binding problem may be resolved, and that such a system readily demonstrates the property of priming. We compare the properties of our system with the general requirements for neural networks that demonstrate compositionality.

Single unit activity in the auditory cortex of a monkey performing a short term memory task

SummaryShort term memory to tones (STMT) was investigated by recording single unit activity in the auditory cortex of a behaving monkey. The activity of each unit was studied in two behavioral conditions: a) During task performance, the monkey had to compare two tones separated by one second of silence (inter-stimulus interval), b) During a nonperforming period; the monkey heard the two tones but did not respond behaviorally. It was noted that the firing rate of many units during the inter-stimulus interval (ISI) was dependent on the frequency of the first tone. Such dependency was observed even towards the end of the ISI, both during task performance trials (50% of the units) and during the nonperforming period (32% of the units). The activity of these units could be the basis of STMT in both of these behavioral states. In 65% of all the units tested, the responses during the ISI were of a higher magnitude in the performance period than were the responses in the non-performance period. The activity of these units may be related either to general processes such as attention and expectation or to short-term memory processes. During task performance, the responses of 23% of the units to the second tone were dependent on whether its frequency was identical to that of the first tone. Such dependency was never observed during the non-performing period. These units may detect similarity or non similarity between two tones presented one second apart. Periodic patterns of firing were not found in the study, thus suggesting that the ISI responses were not generated by reverberatory activity in simple closed loops. On the basis of these results, several alternative mechanisms of STMT are suggested.

Detecting precise firing sequences in experimental data

A precise firing sequence (PFS) is defined here as a sequence of three spikes with fixed delays (up to some time accuracy Delta), that repeat excessively. This paper provides guidelines for detecting PFSs, verifying their significance through surrogate spike trains, and identifying existing PFSs. The method is based on constructing a three-fold correlation among spikes, estimating the expected shape of the correlation by smoothing, and detecting points for which the correlations significantly protrude above the expected correlation. Validation is achieved by generating surrogate spike trains in which the time of each of the real spikes is randomly jittered within a small time window. The method is extensively tested through application to simulated spike trains, and the results are illustrated with recordings of single units in the frontal cortex of behaving monkeys. Pitfalls which may cause false detection of PFSs, or loss of existing PFSs, include searching for PFSs in which the same neuron participates more than once, and attempting to produce a surrogate with some fixed statistical property.

Firing patterns of single units in the prefrontal cortex and neural network models

The occurrences of high-frequency bursts of neural activity and the probability distribution of firing rates of neurons were investigated in the behaving monkey. The activity of 8 to 11 single units was recorded in parallel through six metal microelectrodes from the frontal cortical areas. High-frequency (>150 Hz) bursts of three or more spikes in succession were extremely rare, occurring at an average rate of 0.068 per second per neuron. The probability of observing a burst in one neuron was not affected by the fact that another adjacent neuron emitted a burst. Thus, if a high-frequency burst represents the ‘on’ state of a neuron, we failed to demonstrate persistent states in which a group of neurons is turned ‘on’ together. The firing rates of the neurons were usually low. The probability of a neuron firing a few (2–5) spikes within a narrow (10–100 ms) time window was very low (<0.003). The probability density of firing rates did not show any sign of inhomogeneity, thus failing to show that the neurons...

On embedding synfire chains in a balanced network

We investigate the formation of synfire waves in a balanced network of integrate-and-fire neurons. The synaptic connectivity of this network embodies synfire chains within a sparse random connectivity. This network can exhibit global oscillations but can also operate in an asynchronous activity mode. We analyze the correlations of two neurons in a pool as convenient indicators for the state of the network. We find, using different models, that these indicators depend on a scaling variable. Beyond a critical point, strong correlations and large network oscillations are obtained. We looked for the conditions under which a synfire wave could be propagated on top of an otherwise asynchronous state of the network. This condition was found to be highly restrictive, requiring a large number of neurons for its implementation in our network. The results are based on analytic derivations and simulations.