Article

# Expert-like performance of an autonomous spike tracking algorithm in isolating and maintaining single units in the macaque cortex.

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

## No full-text available

... But stereotrode or tetrode signals need to be combined with 3D spike recognition software tools (e.g., Spike2 from CED, Neuralynx, AutoClass or OpenElectrophy) to reduce the error level from 0–30 % with manual spike sorting to 0–8 % using a semiautomatic spike sorting program [24]. Improper spike sorting may be due to problems identifying single neurons in close proximity to each other when these fire both complex and single spikes (e.g., pyramidal neurons in hippocampus [24] and Purkinje cells in cerebellum [23]), or when spike amplitude varies over time resulting in cluster drift242526. Identification of the type of active neurons is not always done and multiunit activity (MUA) is reported instead. ...
Chapter
Full-text available
In order to assess perfusion and metabolic responses in relation to neural function, the cellular signaling network, including the types of neurons and astrocytes involved, and the timing of their activation need to be known/specified. Here, we present the basis for measuring brain activity and metabolism in rats and mice, which covers basic electrophysiological indicators of neuronal function, a short description of the methods commonly used for recording electrophysiological signals, examples of data analysis and (a brief look at the limitations of the methods. This chapter describes animal preparation, the origin of extracellularly recorded electrical signals, with special regard to the EEG, local field potentials, and spikes (action potentials?) in rodent preparations. We also describe methods for recording cerebral blood flow (CBF), tissue partial pressure of oxygen (tpO2), and cytosolic calcium transients. Lastly, we give examples of protocols in which electrophysiology, blood flow, cerebral rate of oxygen metabolism (CMRO2), and calcium transients have been studied together.
... One source of non-stationarity is the slight drift in the position of the recording electrode relative to the cell body over time which causes gradual changes in the waveform shapes of the recorded spikes ( Lewicki, 1998;Snider and Bonds, 1998). Heartbeat, respiration or mechanical perturbations by movements of the animal in awake preparations, can also cause non-stationarities ( Snider and Bonds, 1998;Chakrabarti et al., 2012). Non-stationary may be caused by a process of electrode encapsulation by fibrous tissue and cell death in the vicinity of the electrode ( Polikov et al., 2005). ...
Article
Objectives Microelectrode arrays offer a means to probe the functional circuitry of the brain and the promise of cortical neuroprosthesis for individuals suffering from paralysis or limb loss. These devices are typically comprised of one or more shanks incorporating microelectrode sites, where the shanks are positioned by inserting the devices along a straight path that is normal to the brain surface. The lack of consistent long‐term chronic recording technology has driven interest in novel probe design and approaches that go beyond the standard insertion approach that is limited to a single velocity or axis. This review offers a description of typical approaches and associated limitations and surveys emergent methods for implantation of microelectrode arrays, in particular those new approaches that leverage embedded microactuators and extend the insertion direction beyond a single axis. Materials and Methods This review paper surveys the current technologies that enable probe implantation, repositioning, and the capability to record/stimulate from a tissue volume. A comprehensive literature search was performed using PubMed, Web of Science, and Google Scholar. Results There has been substantial innovation in the development of microscale and embedded technology that enables probe repositioning to maintain quality recordings in the brain. Innovations in material science have resulted in novel strategies for deployable structures that can record from or stimulate a tissue volume. Moreover, new developments involving magnetically steerable catheters and needles offer an alternative approach to “pull” rather than “push” a probe into the tissue. Conclusion We envision the emergence of a new generation of probes and insertion methodologies for neuromodulation applications that enable reliable chronic performance from devices that can be positioned virtually anywhere in the brain.
Article
Full-text available
Although extracellular unit recording is typically used for the detection of spike occurrences, it also has the theoretical ability to report about what are typically considered intracellular features of the action potential. We address this theoretical ability by developing a model system that captures features of experimentally recorded simultaneous intracellular and extracellular recordings of CA1 pyramidal neurons. We use the line source approximation method of Holt and Koch to model the extracellular action potential (EAP) voltage resulting from the spiking activity of individual neurons. We compare the simultaneous intracellular and extracellular recordings of CA1 pyramidal neurons recorded in vivo with model predictions for the same cells reconstructed and simulated with compartmental models. The model accurately reproduces both the waveform and the amplitude of the EAPs, although it was difficult to achieve simultaneous good matches on both the intracellular and extracellular waveforms. This suggests that accounting for the EAP waveform provides a considerable constraint on the overall model. The developed model explains how and why the waveform varies with electrode position relative to the recorded cell. Interestingly, each cell's dendritic morphology had very little impact on the EAP waveform. The model also demonstrates that the varied composition of ionic currents in different cells is reflected in the features of the EAP.
Article
Full-text available
Neuronal noise sources and systematic variability in the shape of a spike limit the ability to sort multiple unit waveforms recorded from nervous tissue into their single neuron constituents. Here we present a procedure to efficiently sort spikes in the presence of noise that is anisotropic, i.e., dominated by particular frequencies, and whose amplitude distribution may be non-Gaussian, such as occurs when spike waveforms are a function of interspike interval. Our algorithm uses a hierarchical clustering scheme. First, multiple unit records are sorted into an overly large number of clusters by recursive bisection. Second, these clusters are progressively aggregated into a minimal set of putative single units based on both similarities of spike shape as well as the statistics of spike arrival times, such as imposed by the refractory period. We apply the algorithm to waveforms recorded with chronically implanted micro-wire stereotrodes from neocortex of behaving rat. Natural extensions of the algorithm may be used to cluster spike waveforms from records with many input channels, such as those obtained with tetrodes and multiple site optical techniques.
Article
Full-text available
This paper summarizes an algorithm to autonomously position an extracellular recording electrode so as to first isolate the action potentials of a single neuron in a multi-unit signal, and then re-position the electrode as necessary to optimize and maintain the recording quality of that neuron over an extended recording interval. We first summarize some of the technical advancements of the current algorithm over earlier versions of the SpikeTrack recording system in the area of multi-hypothesis cluster tracking method for spike sorting, and a new technique to optimize the signal recording interval. Novel recording experiments in macaque cortex compare the performance of autonomous extracellular recording with that of an experienced neurophysiologist. We found that the algorithm isolates cells better than a human expert.
Conference Paper
Full-text available
This paper introduces a new, unsupervised method for sorting and tracking the non-stationary spike signals of individual neurons in multi-unit extracellular recordings. While this method may be applied to a variety of problems that arise in the field of neural interfaces, its development is motivated by a new class of autonomous neural recording devices. The core of the proposed strategy relies upon an extension of a traditional expectation-maximization (EM) mixture model optimization to incorporate clustering results from the preceding recording interval in a Bayesian manner. Explicit filtering equations for the case of a Gaussian mixture are derived. Techniques using prior data to seed the EM iterations and to select the appropriate model class are also developed. As a natural byproduct of the sorting method, current and prior signal clusters can be matched over time in order to track persisting neurons. Applications of this signal classification method to recordings from macaque parietal cortex show that it provides significantly more consistent clustering and tracking results than traditional methods.
Article
Full-text available
Mounting evidence suggests that understanding how the brain encodes information and performs computations will require studying the correlations between neurons. The recent advent of recording techniques such as multielectrode arrays and two-photon imaging has made it easier to measure correlations, opening the door for detailed exploration of their properties and contributions to cortical processing. However, studies have reported discrepant findings, providing a confusing picture. Here we briefly review these studies and conduct simulations to explore the influence of several experimental and physiological factors on correlation measurements. Differences in response strength, the time window over which spikes are counted, spike sorting conventions and internal states can all markedly affect measured correlations and systematically bias estimates. Given these complicating factors, we offer guidelines for interpreting correlation data and a discussion of how best to evaluate the effect of correlations on cortical processing.
Article
Full-text available
Neurons integrate inputs from thousands of afferents. Similarly, some experimental techniques record the pooled activity of large populations of cells. When cells in these populations are correlated, the correlation coefficient between the collective activity of two subpopulations is typically much larger than the correlation coefficient between individual cells: The act of pooling individual cell signals amplifies correlations. We give an overview of this phenomenon and present several implications. In particular, we show that pooling leads to synchronization in feedforward networks and that it can amplify and otherwise distort correlations between recorded signals.
Article
Full-text available
Studies of neuronal activity throughout the brain predominantly rely on extracellular recordings of spiking activity. These records often contain action potentials from more than one neuron, so it is imperative to discriminate the spikes that originate from separate neurons. The process of discrimination is based on differences in each neuron’s extracellular waveform that originate from cell-specific biophysical properties. We briefly review the electronics of extracellular recording, the detection of spike events, and the clustering of similar spike waveforms as a means to sort spikes according to their putative underlying sources. When spike sorting succeeds, it transforms a fundamental weakness of extracellular recording, namely the inability to isolate single neurons, into one of its greatest strengths, the simultaneous measurement from multiple cells. Special emphasis is placed on visualization schemes for spike data as well as on a set of metrics to estimate the number of incorrectly categorized spikes. False negative contributions to a cluster lead to a suppression of inferred spike rates, while false positive contributions lead to a distortion in the inferred receptive field for the cell. Both errors reduce the estimated information carried by the cell. A matrix of values for these metrics allows readers to assess claims, e.g., the size and reliability of multiple peaks in a receptive field, relative to the level of contamination of the data.
Article
Full-text available
In monkeys, long-term recordings with chronically implanted microelectrodes frequently suffer from a continuously decreasing probability to record single units or even small multiunit clusters. This problem is associated with two technical limitations of the available devices: first, restrictions for electrode movement, and second, absent possibility to exchange electrodes easily on a regular basis. Permitting to adjust the recording site and to use new recording tracks with proper electrodes may avoid these problems and make chronic more similar to acute recordings. Here, we describe a novel type of implant tackling this issue. It consists of a new type of recording chamber combined with an exchangeable multielectrode array that precisely fits into it. The multielectrode array is reversibly fixed to the chamber, and within a minute it can be exchanged against another array equipped with new electrodes at the awake animal. The array allows for bidirectional movement of six electrodes for a distance of up to 12 mm. The recording chamber enables hermetical isolation of the intracranial space, resulting in long-lasting aseptic conditions and reducing dural thickening to a minimum, as confirmed by microbiological and histopathological analysis. The device has a simple design and is both easy to produce and low in cost. Functionality has been tested in primary and secondary visual cortex of three macaque monkeys over a period of up to 15 mo. The results show that even after more than a year, single and multiunit responses can be obtained with high incidence.
Article
Full-text available
A prominent feature of motor cortex field potentials during movement is a distinctive low-frequency local field potential (lf-LFP) (<4 Hz), referred to as the movement event-related potential (mEP). The lf-LFP appears to be a global signal related to regional synaptic input, but its relationship to nearby output signaled by single unit spiking activity (SUA) or to movement remains to be established. Previous studies comparing information in primary motor cortex (MI) lf-LFPs and SUA in the context of planar reaching tasks concluded that lf-LFPs have more information than spikes about movement. However, the relative performance of these signals was based on a small number of simultaneously recorded channels and units, or for data averaged across sessions, which could miss information of larger-scale spiking populations. Here, we simultaneously recorded LFPs and SUA from two 96-microelectrode arrays implanted in two major motor cortical areas, MI and ventral premotor (PMv), while monkeys freely reached for and grasped objects swinging in front of them. We compared arm end point and grip aperture kinematics' decoding accuracy for lf-LFP and SUA ensembles. The results show that lf-LFPs provide enough information to reconstruct kinematics in both areas with little difference in decoding performance between MI and PMv. Individual lf-LFP channels often provided more accurate decoding of single kinematic variables than any one single unit. However, the decoding performance of the best single unit among the large population usually exceeded that of the best single lf-LFP channel. Furthermore, ensembles of SUA outperformed the pool of lf-LFP channels, in disagreement with the previously reported superiority of lf-LFP decoding. Decoding results suggest that information in lf-LFPs recorded from intracortical arrays may allow the reconstruction of reach and grasp for real-time neuroprosthetic applications, thus potentially supplementing the ability to decode these same features from spiking populations.
Article
Full-text available
One of the critical requirements of the emerging class of neural prosthetic devices is to maintain good quality neural recordings over long time periods. We report here a novel MEMS (Micro Electro Mechanical Systems) based technology that can move microelectrodes in the event of deterioration in neural signal to sample a new set of neurons. Microscale electro-thermal actuators are used to controllably move microelectrodes post-implantation in steps of approximately 9 mum. In this study, a total of 12 movable microelectrode chips were individually implanted in adult rats. Two of the twelve movable microelectrode chips were not moved over a period of 3 weeks and were treated as control experiments. During the first 3 weeks of implantation, moving the microelectrodes led to an improvement in the average signal to noise ratio (SNR) from 14.61 +/- 5.21 dB before movement to 18.13 +/- 4.99 dB after movement across all microelectrodes and all days. However, the average root-mean-square values of noise amplitudes were similar at 2.98 +/- 1.22 muV and 3.01 +/- 1.16 muV before and after microelectrode movement. Beyond 3 weeks, the primary observed failure mode was biological rejection of the PMMA (dental cement) based skull mount resulting in the device loosening and eventually falling from the skull. Additionally, the average SNR for functioning devices beyond 3 weeks was 11.88 +/- 2.02 dB before microelectrode movement and was significantly different (p < 0.01) from the average SNR of 13.34 +/- 0.919 dB after movement. The results of this study demonstrate that MEMS based technologies can move microelectrodes in rodent brains in long-term experiments resulting in improvements in signal quality. Further improvements in packaging and surgical techniques will potentially enable movable microelectrodes to record cortical neuronal activity in chronic experiments.
Article
Full-text available
In this paper we describe a set of algorithms and a novel miniature device that together can autonomously position electrodes in neural tissue to obtain high-quality extracellular recordings. This robotic system moves each electrode to detect the signals of individual neurons, optimize the signal quality of a target neuron, and then maintain this signal over time. Such neuronal signals provide the key inputs for emerging neuroprosthetic medical devices and serve as the foundation of basic neuroscientific and medical research. Experimental results from extensive use of the robotic electrodes in macaque parietal cortex are presented to validate the method and to quantify its effectiveness.
Article
Full-text available
We simulated the shape and amplitude of extracellular action potentials (APs or "spikes") using biophysical models based on detailed reconstructions of single neurons from the cat's visual cortex. We compared these predictions with spikes recorded from the cat's primary visual cortex under a standard protocol. The experimental data were derived from a large number of neurons throughout all layers. The majority of spikes were biphasic, with a dominant negative peak (mean amplitude, -0.11 mV), whereas a minority of APs had a dominant positive peak of +0.54-mV mean amplitude, with a maximum of +1.5 mV. The largest positive amplitude spikes were recorded in layer 5. The simulations demonstrated that a pyramidal neuron under known biophysical conditions may generate a negative peak with amplitude up to -1.5 mV, but that the amplitude of the positive peak may be at most 0.5 mV. We confirmed that spikes with large positive peaks were not produced by juxtacellular patch recordings. We conclude that there is a significant gap in our present understanding of either the spike-generation process in pyramidal neurons, the biophysics of extracellular recording, or both.
Article
Full-text available
Cortical control of neuroprosthetic devices is known to require neuronal adaptations. It remains unclear whether a stable cortical representation for prosthetic function can be stored and recalled in a manner that mimics our natural recall of motor skills. Especially in light of the mixed evidence for a stationary neuron-behavior relationship in cortical motor areas, understanding this relationship during long-term neuroprosthetic control can elucidate principles of neural plasticity as well as improve prosthetic function. Here, we paired stable recordings from ensembles of primary motor cortex neurons in macaque monkeys with a constant decoder that transforms neural activity to prosthetic movements. Proficient control was closely linked to the emergence of a surprisingly stable pattern of ensemble activity, indicating that the motor cortex can consolidate a neural representation for prosthetic control in the presence of a constant decoder. The importance of such a cortical map was evident in that small perturbations to either the size of the neural ensemble or to the decoder could reversibly disrupt function. Moreover, once a cortical map became consolidated, a second map could be learned and stored. Thus, long-term use of a neuroprosthetic device is associated with the formation of a cortical map for prosthetic function that is stable across time, readily recalled, resistant to interference, and resembles a putative memory engram.
Article
Full-text available
The ultimate goal of neural interface research is to create links between the nervous system and the outside world either by stimulating or by recording from neural tissue to treat or assist people with sensory, motor, or other disabilities of neural function. Although electrical stimulation systems have already reached widespread clinical application, neural interfaces that record neural signals to decipher movement intentions are only now beginning to develop into clinically viable systems to help paralyzed people. We begin by reviewing state-of-the-art research and early-stage clinical recording systems and focus on systems that record single-unit action potentials. We then address the potential for neural interface research to enhance basic scientific understanding of brain function by offering unique insights in neural coding and representation, plasticity, brain-behavior relations, and the neurobiology of disease. Finally, we discuss technical and scientific challenges faced by these systems before they are widely adopted by severely motor-disabled patients.
Article
Full-text available
Multiple single-unit recording has become one of the most powerful in vivo electro-physiological techniques for studying neural circuits. The demand has been increasing for small and lightweight chronic recording devices that allow fine adjustments to be made over large numbers of electrodes across multiple brain regions. To achieve this, we developed precision motorized microdrive arrays that use a novel motor multiplexing headstage to dramatically reduce wiring while preserving precision of the microdrive control. Versions of the microdrive array were chronically implanted on both rats (21 microdrives) and mice (7 microdrives), and relatively long-term recordings were taken.
Article
Full-text available
It is now commonly accepted that planning and execution of movements are based on distributed processing by neuronal populations in motor cortical areas. It is less clear, though, how these populations organize dynamically to cope with the momentary computational demands. Simultaneously recorded activities of neurons in the primary motor cortex of monkeys during performance of a delayed-pointing task exhibited context-dependent, rapid changes in the patterns of coincident action potentials. Accurate spike synchronization occurred in relation to external events (stimuli, movements) and was commonly accompanied by discharge rate modulations but without precise time locking of the spikes to these external events. Spike synchronization also occurred in relation to purely internal events (stimulus expectancy), where firing rate modulations were distinctly absent. These findings indicate that internally generated synchronization of individual spike discharges may subserve the cortical organization of cognitive motor processes.
Article
Full-text available
The detection of neural spike activity is a technical challenge that is a prerequisite for studying many types of brain function. Measuring the activity of individual neurons accurately can be difficult due to large amounts of background noise and the difficulty in distinguishing the action potentials of one neuron from those of others in the local area. This article reviews algorithms and methods for detecting and classifying action potentials, a problem commonly referred to as spike sorting. The article first discusses the challenges of measuring neural activity and the basic issues of signal detection and classification. It reviews and illustrates algorithms and techniques that have been applied to many of the problems in spike sorting and discusses the advantages and limitations of each and the applicability of these methods for different types of experimental demands. The article is written both for the physiologist wanting to use simple methods that will improve experimental yield and minimize the selection biases of traditional techniques and for those who want to apply or extend more sophisticated algorithms to meet new experimental challenges.
Article
Full-text available
To determine whether simultaneously recorded motor cortex neurons can be used for real-time device control, rats were trained to position a robot arm to obtain water by pressing a lever. Mathematical transformations, including neural networks, converted multineuron signals into 'neuronal population functions' that accurately predicted lever trajectory. Next, these functions were electronically converted into real-time signals for robot arm control. After switching to this 'neurorobotic' mode, 4 of 6 animals (those with > 25 task-related neurons) routinely used these brain-derived signals to position the robot arm and obtain water. With continued training in neurorobotic mode, the animals' lever movement diminished or stopped. These results suggest a possible means for movement restoration in paralysis patients.
Article
Full-text available
Recent technological and scientific advances have generated wide interest in the possibility of creating a brain-machine interface (BMI), particularly as a means to aid paralyzed humans in communication. Advances have been made in detecting neural signals and translating them into command signals that can control devices. We now have systems that use externally derived neural signals as a command source, and faster and potentially more flexible systems that directly use intracortical recording are being tested. Studies in behaving monkeys show that neural output from the motor cortex can be used to control computer cursors almost as effectively as a natural hand would carry out the task. Additional research findings explore the possibility of using computers to return behaviorally useful feedback information to the cortex. Although significant scientific and technological challenges remain, progress in creating useful human BMIs is accelerating.
Article
Full-text available
We have recently developed a closed-loop environment in which we can test the ability of primates to control the motion of a virtual device using ensembles of simultaneously recorded neurons /29/. Here we use a maximum likelihood method to assess the information about task performance contained in the neuronal ensemble. We trained two animals to control the motion of a computer cursor in three dimensions. Initially the animals controlled cursor motion using arm movements, but eventually they learned to drive the cursor directly from cortical activity. Using a population vector (PV) based upon the relation between cortical activity and arm motion, the animals were able to control the cursor directly from the brain in a closed-loop environment, but with difficulty. We added a supervised learning method that modified the parameters of the PV according to task performance (adaptive PV), and found that animals were able to exert much finer control over the cursor motion from brain signals. Here we describe a maximum likelihood method (ML) to assess the information about target contained in neuronal ensemble activity. Using this method, we compared the information about target contained in the ensemble during arm control, during brain control early in the adaptive PV, and during brain control after the adaptive PV had settled and the animal could drive the cursor reliably and with fine gradations. During the arm-control task, the ML was able to determine the target of the movement in as few as 10% of the trials, and as many as 75% of the trials, with an average of 65%. This average dropped when the animals used a population vector to control motion of the cursor. On average we could determine the target in around 35% of the trials. This low percentage was also reflected in poor control of the cursor, so that the animal was unable to reach the target in a large percentage of trials. Supervised adjustment of the population vector parameters produced new weighting coefficients and directional tuning parameters for many neurons. This produced a much better performance of the brain-controlled cursor motion. It was also reflected in the maximum likelihood measure of cell activity, producing the correct target based only on neuronal activity in over 80% of the trials on average. The changes in maximum likelihood estimates of target location based on ensemble firing show that an animal's ability to regulate the motion of a cortically controlled device is not crucially dependent on the experimenter's ability to estimate intention from neuronal activity.
Article
Full-text available
A paradigm is described for recording the activity of single cortical neurons from awake, behaving macaque monkeys. Its unique features include high-density microwire arrays and multichannel instrumentation. Three adult rhesus monkeys received microwire array implants, totaling 96-704 microwires per subject, in up to five cortical areas, sometimes bilaterally. Recordings 3-4 weeks after implantation yielded 421 single neurons with a mean peak-to-peak voltage of 115 +/- 3 microV and a signal-to-noise ratio of better than 5:1. As many as 247 cortical neurons were recorded in one session, and at least 58 neurons were isolated from one subject 18 months after implantation. This method should benefit neurophysiological investigation of learning, perception, and sensorimotor integration in primates and the development of neuroprosthetic devices.
Article
Full-text available
Control of prostheses using cortical signals is based on three elements: chronic microelectrode arrays, extraction algorithms, and prosthetic effectors. Arrays of microelectrodes are permanently implanted in cerebral cortex. These arrays must record populations of single- and multiunit activity indefinitely. Information containing position and velocity correlates of animate movement needs to be extracted continuously in real time from the recorded activity. Prosthetic arms, the current effectors used in this work, need to have the agility and configuration of natural arms. Demonstrations using closed-loop control show that subjects change their neural activity to improve performance with these devices. Adaptive-learning algorithms that capitalize on these improvements show that this technology has the capability of restoring much of the arm movement lost with immobilizing deficits.
Article
Full-text available
A system was developed that can autonomously position recording electrodes to isolate and maintain optimal quality extracellular signals. The system consists of a novel motorized miniature recording microdrive and a control algorithm. The microdrive was designed for chronic operation and can independently position four glass-coated Pt-Ir electrodes with micrometer precision over a 5-mm range using small (3 mm diam) piezoelectric linear actuators. The autonomous positioning algorithm is designed to detect, align, and cluster action potentials and then command the microdrive to optimize and maintain the neural signal. This system is shown to be capable of autonomous operation in monkey cortical tissue.
Article
Full-text available
Recent development of neural prosthetics for assisting paralyzed patients has focused on decoding intended hand trajectories from motor cortical neurons and using this signal to control external devices. In this study, higher level signals related to the goals of movements were decoded from three monkeys and used to position cursors on a computer screen without the animals emitting any behavior. Their performance in this task improved over a period of weeks. Expected value signals related to fluid preference, the expected magnitude, or probability of reward were decoded simultaneously with the intended goal. For neural prosthetic applications, the goal signals can be used to operate computers, robots, and vehicles, whereas the expected value signals can be used to continuously monitor a paralyzed patient's preferences and motivation.
Article
We discuss the use of standard logistic regression techniques to estimate hazard rates and survival curves from censored data. These techniques allow the statistician to use parametric regression modeling on censored data in a flexible way that provides both estimates and standard errors. An example is given that demonstrates the increased structure that can be seen in a parametric analysis, as compared with the nonparametric Kaplan-Meier survival curves. In fact, the logistic regression estimates are closely related to Kaplan-Meier curves, and approach the Kaplan-Meier estimate as the number of parameters grows large.
Article
We show that the standard filters used for on-line spike detection in most hardware acquisition systems introduce distortions in the recorded spike shapes. This is because on-line spike detection is done after band pass filtering the data with causal filters. As illustrated with three clusters of spike shapes from a real single cell recording in a human subject, causal filtering can create a spurious negative rebound and a smooth looking appearance of the spikes. We also show that these filtering distortions can make artifacts look similar to real spikes.
Article
The use of chronically implanted electrodes for neural recordings in small, freely behaving animals poses several unique technical challenges. Because of the need for an extremely lightweight apparatus, chronic recording technology has been limited to manually operated microdrives, despite the advantage of motorized manipulators for positioning electrodes. Here we describe a motorized, miniature chronically implantable microdrive for independently positioning three electrodes in the brain. The electrodes are controlled remotely, avoiding the need to disturb the animal during electrode positioning. The microdrive is approximately 6 mm in diameter, 17 mm high and weighs only 1.5 g, including the headstage preamplifier. Use of the motorized microdrive has produced a ten-fold increase in our data yield compared to those experiments done using a manually operated drive. In addition, we are able to record from multiple single neurons in the behaving animal with signal quality comparable to that seen in a head-fixed anesthetized animal. We also describe a motorized commutator that actively tracks animal rotation based on a measurement of torque in the tether.
Conference Paper
This paper introduces an algorithm for tracking targets whose locations are inferred from clusters of observations. This method, which we call MHTC, expands the traditional multiple hypothesis tracking (MHT) hypothesis tree to include model hypotheses - possible ways the data can be clustered in each time step - as well as ways the measurements can be associated with existing targets across time steps. We present this new hypothesis framework and its probability expressions and demonstrate MHTC's operation in a robotic solution to tracking neural signal sources.
Article
In lifetesting, medical follow-up, and other fields the observation of the time of occurrence of the event of interest (called a death) may be prevented for some of the items of the sample by the previous occurrence of some other event (called a loss). Losses may be either accidental or controlled, the latter resulting from a decision to terminate certain observations. In either case it is usually assumed in this paper that the lifetime (age at death) is independent of the potential loss time; in practice this assumption deserves careful scrutiny. Despite the resulting incompleteness of the data, it is desired to estimate the proportion P(t) of items in the population whose lifetimes would exceed t (in the absence of such losses), without making any assumption about the form of the function P(t). The observation for each item of a suitable initial event, marking the beginning of its lifetime, is presupposed. For random samples of size N the product-limit (PL) estimate can be defined as follows: List and label the N observed lifetimes (whether to death or loss) in order of increasing magnitude, so that one has $$0 \leqslant t_1^\prime \leqslant t_2^\prime \leqslant \cdots \leqslant t_N^\prime .$$ Then $$\hat P\left( t \right) = \Pi r\left[ {\left( {N - r} \right)/\left( {N - r + 1} \right)} \right]$$, where r assumes those values for which $$t_r^\prime \leqslant t$$ and for which $$t_r^\prime$$ measures the time to death. This estimate is the distribution, unrestricted as to form, which maximizes the likelihood of the observations. Other estimates that are discussed are the actuarial estimates (which are also products, but with the number of factors usually reduced by grouping); and reduced-sample (RS) estimates, which require that losses not be accidental, so that the limits of observation (potential loss times) are known even for those items whose deaths are observed. When no losses occur at ages less than t the estimate of P(t) in all cases reduces to the usual binomial estimate, namely, the observed proportion of survivors.
Article
Variability of single-unit neural recordings can significantly affect the overall performance achieved by brain machine interfaces (BMI). In this paper, we present a novel technique to adapt a linear filter commonly used in BMI to compensate for loss of neurons from the recorded neural ensemble, thus minimizing loss in performance. We simulate the gains achieved by this technique using a model of the learning process during closed-loop BMI operation. This simulation suggests that we can adapt to the loss of 24% of the neurons controlling a BMI with only 13% drop in performance.
Article
Multielectrode recording experiments let us measure correlations between the activity of individual neurons and the neural circuits in which they are embedded. Recently, multielectrode studies have been emphasizing how correlated neuronal activity is linked with behavior. Decisions are fundamental to voluntary behavior. Here, we discuss computations necessary to turn a decision into an action and review progress in studying correlated neural activity in areas of the brain which link sensory and motor representations. The themes that emerge are that correlated patterns of activity in populations of neurons can be revealed by measurements of field potential fluctuations and that these measurements can relate the activity of individual neurons to the activity of populations of neurons distributed across different regions of the brain. Investigations into patterns of neuronal correlation are helping us to understand how decisions and other cognitive processes result from the interactions between different brain systems that are responsible for controlling and regulating our behavior.
Article
This paper introduces a new, unsupervised method for sorting and tracking the action potentials of individual neurons in multiunit extracellular recordings. Presuming the data are divided into short, sequential recording intervals, the core of our strategy relies upon an extension of a traditional mixture model approach that incorporates clustering results from the preceding interval in a Bayesian manner, while still allowing for signal nonstationarity and changing numbers of recorded neurons. As a natural byproduct of the sorting method, current and prior signal clusters can be matched over time in order to track persisting neurons. We also develop techniques to use prior data to appropriately seed the clustering algorithm and select the model class. We present results in a principal components space; however, the algorithm may be applied in any feature space where the distribution of a neuron's spikes may be modeled as Gaussian. Applications of this signal classification method to recordings from macaque parietal cortex show that it provides significantly more consistent clustering and tracking results than traditional methods based on expectation–maximization optimization of mixture models. This consistent tracking ability is crucial for intended applications of the method.
Article
Neural interfaces (NIs) for motor control have recently become increasingly advanced. This has been possible owing to substantial progress in our understanding of the cortical motor system as well as the development of appropriate decoding methods in both non-human primates and paralyzed patients. So far, neural interfaces have controlled mainly computer screens and robotic arms. An important advancement has been the demonstration of neural interfaces that can directly control the subject's muscles. Furthermore, it has been shown that cortical plasticity alone can optimize neural interface performance in the absence of machine learning, which emphasizes the role of the brain for neural interface adaptation. Future motor prostheses may use also sensory feedback to enhance their control capabilities.
Article
The spiking activity of cortical neurons is correlated. For instance, trial-to-trial fluctuations in response strength are shared between neurons, and spikes often occur synchronously. Understanding the properties and mechanisms that generate these forms of correlation is critical for determining their role in cortical processing. We therefore investigated the spatial extent and functional specificity of correlated spontaneous and evoked activity. Because feedforward, recurrent, and feedback pathways have distinct extents and specificity, we reasoned that these measurements could elucidate the contribution of each type of input. We recorded single unit activity with microelectrode arrays which allowed us to measure correlation in many hundreds of pairings, across a large range of spatial scales. Our data show that correlated evoked activity is generated by two mechanisms that link neurons with similar orientation preferences on different spatial scales: one with high temporal precision and a limited spatial extent (approximately 3 mm), and a second that gives rise to correlation on a slow time scale and extends as far as we were able to measure (10 mm). The former is consistent with common input provided by horizontal connections; the latter likely involves feedback from extrastriate cortex. Spontaneous activity was correlated over a similar spatial extent, but approximately twice as strongly as evoked activity. Visual stimuli thus caused a substantial decrease in correlation, particularly at response onset. These properties and the circuit mechanism they imply provide new constraints on the functional role that correlation may play in visual processing.
Article
1. A particular firing pattern among simultaneously observed neurons represents a particular sequence of activity. If any multineuron pattern repeats significantly more than expected by chance, we may be observing a repeated state of a neural assembly as it processes similar units of information. 2. We present here an algorithm that rapidly finds all single or multineuron patterns that repeat two or more times within a block of data, as well as equations for calculating the number of patterns of given length and repetition that would be expected. The complexity of patterns for which it is practical to compute expected numbers is three to six spikes (inclusive). 3. Confidence limits are based on these expected numbers of patterns, so that is possible to identify groups of patterns that are worthy of further analysis. 4. These methods are tested against simulated multineuron data that has various types of known nonstationarities, with good agreement between observed and expected values. 5. Application to real spike trains shows a large excess of observed repeating patterns, of which some, but not all, are shown to be due to bursts of high frequency firing. 6. It should be possible to apply the new method as a filter in real time in order to search for an association between repeated pattern events and externally observable events (stimulus, behavior, etc.). Any repeated pattern events which cannot be so associated may represent a new indicator of internal events in the nervous system.
Article
We developed a new method for the insertion of thin-shaft probes into neural and muscular tissue. Axial forces for driving the probes into tissue and radial forces against buckling are both provided by a stretched elastic rubber tube in which the probe is guided outside the tissue. Various geometric arrangements of arrays with independently advanceable probes are possible. Prototypes with 7 linearly aligned fiber electrodes and computer-controlled positioning motors were successfully used in single- and multiple-unit recordings from the visual system of awake monkeys (Eckhorn et al., 1993). The method is suitable in a wide range of applications, including insertion of fine microprobe fibers and wire electrodes into brain and muscle through the skin or dura, provided that the tips of the probes are sharp and hard enough.
Article
Although a number of methods have been proposed for classification of individual action potentials embedded in multi-unit activity, they have been challenged by non-stationarity. The waveform shapes of action potentials can change rapidly over time as a result of shifts in membrane conductances during extended burst firing sequences and more slowly over time due to electrode drift. These changes are typically non-Gaussian. We present an algorithm for waveform identification that makes no assumptions on the distribution of these shapes other than the change in waveform shape for a particular neuron should not be discontinuous. We apply this algorithm to the resolution of multi-unit neural signals recorded in the cat visual cortex and we compare this approach to a spike sorting method that is based on the Bayesian likelihood of a spike fitting a particular model (Lewicki, M. Bayesian modeling and classification of neural signals. Neural Comput 1994;6(5):1005-1030.
Article
Ensemble recording in cerebellar cortex of awake rats presents unique methodological challenges not encountered when recording from the cerebral cortex or from deep brain structures with more homogeneous cell populations. Compared to the cerebral cortex, removal of dura over the cerebellum evokes pronounced swelling, and insertion of multiple closely spaced electrodes in the cerebellar cortex causes considerable dimpling (Welsh JP, Schwartz C. Multielectrode recording from the cerebellum. In: Nicolelis MAL, editor. Methods for Neural Ensemble Recordings, CRC Methods in Neuroscience Series. Boca Raton, FL: CRC Press LLC, 1999, pp. 79-100). Also, a repetitious and well-defined neural circuit characterizes the cerebellar cortex across its entire surface. With conventional multi-electrode methods, such as chronically implanted bundles or arrays of microwires, the risk of disrupting the cerebellar cytoarchitecture is high. In most conventional multi-electrode systems, electrodes have rather low impedance and cannot be moved independently after implantation. These limitations make proper unit isolation, necessary to identify each of the recorded cerebellar units, very difficult. We designed a lightweight (14 g), miniature (base plate: 19 x 23 mm; total height: 16 mm) multi-electrode system to allow for the chronic implantation of six independently moveable sharp electrodes with high impedance, in the cerebellar cortex. The six electrodes are arranged in a 2 x 3 matrix (inter-electrode distance: 0.6 mm). At any time after the implantation the vertical position of each individual electrode can be adjusted by screwing spring-loaded electrode heads up or down. The system preserves the integrity of the cerebellar cytoarchitecture, and enables easy isolation and identification of individual cerebellar units in awake, freely moving rats.
Article
Simultaneous recording from multiple single neurones presents many technical difficulties. However, obtaining such data has many advantages, which make it highly worthwhile to overcome the technical problems. This report describes methods which we have developed to permit recordings in awake behaving monkeys using the 'Eckhorn' 16 electrode microdrive. Structural magnetic resonance images are collected to guide electrode placement. Head fixation is achieved using a specially designed headpiece, modified for the multiple electrode approach, and access to the cortex is provided via a novel recording chamber. Growth of scar tissue over the exposed dura mater is reduced using an anti-mitotic compound. Control of the microdrive is achieved by a computerised system which permits several experimenters to move different electrodes simultaneously, considerably reducing the load on an individual operator. Neurones are identified as pyramidal tract neurones by antidromic stimulation through chronically implanted electrodes; stimulus control is integrated into the computerised system. Finally, analysis of multiple single unit recordings requires accurate methods to correct for non-stationarity in unit firing. A novel technique for such correction is discussed.
Article
Real-time direct interfaces between the brain and electronic and mechanical devices could one day be used to restore sensory and motor functions lost through injury or disease. Hybrid brain-machine interfaces also have the potential to enhance our perceptual, motor and cognitive capabilities by revolutionizing the way we use computers and interact with remote environments.
Article
As long as 150 years ago, when Fritz and Hitzig demonstrated the electrical excitability of the motor cortex, scientists and fiction writers were considering the possibility of interfacing a machine with the human brain. Modern attempts have been driven by concrete technological and clinical goals. The most advanced of these has brought the perception of sound to thousands of deaf individuals by means of electrodes implanted in the cochlea. Similar attempts are underway to provide images to the visual cortex and to allow the brains of paralyzed patients to re-establish control of the external environment via recording electrodes. This review focuses on two challenges: (1) establishing a 'closed loop' between sensory input and motor output and (2) controlling neural plasticity to achieve the desired behavior of the brain-machine system. Meeting these challenges is the key to extending the impact of the brain-machine interface.
Article
Micro-machined neural prosthetic devices can be designed and fabricated to permit recording and stimulation of specific sites in the nervous system. Unfortunately, the long-term use of these devices is compromised by cellular encapsulation. The goals of this study were to determine if device size, surface characteristics, or insertion method affected this response. Devices with two general designs were used. One group had chisel-shaped tips, sharp angular corners, and surface irregularities on the micrometer size scale. The second group had rounded corners, and smooth surfaces. Devices of the first group were inserted using a microprocessor-controlled inserter. Devices of the second group were inserted by hand. Comparisons were made of responses to the larger devices in the first group with devices from the second group. Responses were assessed 1 day and 1, 2, 4, 6, and 12 weeks after insertions. Tissues were immunochemically labeled for glial fibrillary acidic protein (GFAP) or vimentin to identify astrocytes, or for ED1 to identify microglia. For the second comparison devices from the first group with different cross-sectional areas were analyzed. Similar reactive responses were observed following insertion of all devices; however, the volume of tissue involved at early times, <1 week, was proportional to the cross-sectional area of the devices. Responses observed after 4 weeks were similar for all devices. Thus, the continued presence of devices promotes formation of a sheath composed partly of reactive astrocytes and microglia. Both GFAP-positive and -negative cells were adherent to all devices. These data indicate that device insertion promotes two responses-an early response that is proportional to device size and a sustained response that is independent of device size, geometry, and surface roughness. The early response may be associated with the amount of damage generated during insertion. The sustained response is more likely due to tissue-device interactions.
Article
METAL microelectrodes suitable for recording from single cells externally or after penetration have been described by various authors1-3. The procedures of Svaetichin2 and Grundfest1 call for special apparatus, and the method of Hubel3, employing tungsten, in our hands gives rather noisy electrodes which are difficult to straighten. Steel electrodes, which have the advantages of low noise, low electrode potential, ease of straightening and, in particular, offer the possibility of accurate localization of the recording site, may be made as follows.
Article
The implantation of chronic recording electrodes in the brain has been shown to be a valuable method for simultaneously recording from many neurons. However, precise placement of these electrodes, crucial for successful recording, is challenging if the target area is not on the brain surface. Here we present a stereotaxic implantation procedure to chronically implant bundles of recording electrodes into macaque cortical sulci, employing magnetic resonance (MR) imaging to determine stereotaxic coordinates of target location and sulcus orientation. Using this method in four animals, we recorded simultaneously the spiking activity and the local field potential from the parietal reach region (PRR), located in the medial bank of the intraparietal sulcus (IPS), while the animal performed a reach movement task. Fifty percent of all electrodes recorded spiking activity during the first 2 post-operative months, indicating their placement within cortical gray matter. Chronic neural activity was similar to standard single electrode recordings in PRR, as reported previously. These results indicate that this MR image-guided implantation technique can provide sufficient placement accuracy in cortical sulci and subcortical structures. Moreover, this technique may be useful for future cortical prosthesis applications in humans that require implants within sulci.
Article
Correlations among simultaneously recorded signals are mostly analyzed pairwise and include temporal averaging. However, pairwise methods are not suitable for characterizing relationships among multiple channels for signals which vary temporally in an unpredictable way. Here we develop a time-resolved spatio-temporal correlation (STC) measure among simultaneously recorded signals. We demonstrate the capabilities of the method with artificial data sets and with multiple-channel recordings from striate cortex of awake monkeys. We concentrate on correlations in the gamma-frequency range (gamma: 30-90 Hz) because they were prominent in the analyzed recordings and gained high interest in the recent years due to their assumed role in associative processing, including perceptual binding. Former analyses of gamma-activities in visual cortex, using pairwise correlation methods, mostly revealed zero-delay correlation, indicating synchrony. In cat and monkey visual cortex this gamma-synchrony is restricted to 1.5-3.0 mm (half-height decline). However, our spatio-temporal correlation (STC)-method demonstrates for striate cortex from awake monkeys that gamma-synchrony is a local phenomenon of more global traveling plane waves that appear stimulus-induced at randomly varying orientations. These gamma-waves are coupled over much larger cortical distances (approximately 7 mm half-height decline) than the gamma-synchrony ranges obtained by pairwise correlation analyses from the same data. Our STC-method therefore suggests that the previously reported results of short-range and zero-delay correlations were often due to temporal averaging of traveling gamma-waves.
Article
Implantable silicon microelectrode array technology is a useful technique for obtaining high-density, high-spatial resolution sampling of neuronal activity within the brain and holds promise for a wide range of neuroprosthetic applications. One of the limitations of the current technology is inconsistent performance in long-term applications. Although the brain tissue response is believed to be a major cause of performance degradation, the precise mechanisms that lead to failure of recordings are unknown. We observed persistent ED1 immunoreactivity around implanted silicon microelectrode arrays implanted in adult rat cortex that was accompanied by a significant reduction in nerve fiber density and nerve cell bodies in the tissue immediately surrounding the implanted silicon microelectrode arrays. Persistent ED1 up-regulation and neuronal loss was not observed in microelectrode stab controls indicating that the phenotype did not result from the initial mechanical trauma of electrode implantation, but was associated with the foreign body response. In addition, we found that explanted electrodes were covered with ED1/MAC-1 immunoreactive cells and that the cells released MCP-1 and TNF-alpha under serum-free conditions in vitro. Our findings suggest a potential new mechanism for chronic recording failure that involves neuronal cell loss, which we speculate is caused by chronic inflammation at the microelectrode brain tissue interface.