Temporal patterning of saccadic eye movement signals

Eye Movements as a Window into Decision-Making

For over 100 years, eye movements have been studied and used as indicators of human sensory and cognitive functions. This review evaluates how eye movements contribute to our understanding of the processes that underlie decision-making. Eye movement metrics signify the visual and task contexts in which information is accumulated and weighed. They indicate the efficiency with which we evaluate the instructions for decision tasks, the timing and duration of decision formation, the expected reward associated with a decision, the accuracy of the decision outcome, and our ability to predict and feel confident about a decision. Because of their continuous nature, eye movements provide an exciting opportunity to probe decision processes noninvasively in real time. Expected final online publication date for the Annual Review of Vision Science, Volume 8 is September 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Population temporal structure supplements the rate code during sensorimotor transformations

Sensorimotor transformations are mediated by premotor brain networks where individual neurons represent sensory, cognitive, and movement-related information. Such multiplexing poses a conundrum-how does a decoder know precisely when to initiate a movement if its inputs are active at times when a movement is not desired (e.g., in response to sensory stimulation)? Here, we propose a novel hypothesis: movement is triggered not only by an increase in firing rate but, critically, also by a reliable temporal pattern in the population response. Laminar recordings in the macaque superior colliculus (SC), a midbrain hub of orienting control, and pseudo-population analyses in SC and cortical frontal eye fields (FEFs) corroborated this hypothesis. Specifically, using a measure that captures the fidelity of the population code-here called temporal stability-we show that the temporal structure fluctuates during the visual response but becomes increasingly stable during the movement command. Importantly, we used spatiotemporally patterned microstimulation to causally test the contribution of population temporal stability in gating movement initiation and found that stable stimulation patterns were more likely to evoke a movement. Finally, a spiking neuron model was able to discriminate between stable and unstable input patterns, providing a putative biophysical mechanism for decoding temporal structure. These findings offer new insights into the long-standing debate on motor preparation and generation by situating the movement gating signal in temporal features of activity in shared neural substrates, and they highlight the importance of short-term population history in neuronal communication and behavior.

Hematology and Culture Assessment of Cranially Implanted Rhesus Macaques (Macaca mulatta)

The use of percutaneous cranial implants in rhesus macaques (Macaca mulatta) has long been a valuable tool for neuroscience research. However, when treating and assessing these animals, veterinarians are required to make assumptions about diagnostic results due to a lack of research into how these implants affect physiology. Microbial cultures of cranial implant sites show an abundance of colonizing bacteria, but whether these microbes affect animal health and wellbeing is poorly understood. In addition, microbial antibiotic resistance can present significant health concerns for both the animals and the researchers. To help elucidate the relationship between percutaneous cranial implants and blood parameters, complete blood cell counts and serum chemistry results were assessed on 57 nonhuman primates at our institution from September 2001 to March 2017. Generalized estimating equations were used to compare the results before and after an animal's first implant surgery. This modelling showed that cranial implants were a significant predictor of alterations in the number of neutrophils, lymphocytes, and red blood cells, and in the concentration of hemoglobin, alkaline phosphatase, creatinine, calcium, phos- phorus, total protein, albumin, and globulin. Anaerobic and aerobic bacterial cultures were performed to identify bacteria associated with cranial implants. Staphylococcus spp., Streptococcus spp., and Corynebacterium spp. comprised the majority of the aerobic bacterial isolates, while Fusobacterium spp., Peptostreptococcus spp. and Bacterioides fragilis comprised the majority of anaerobic bacterial isolates. Using a Pearson r correlation for statistical analysis, we assessed whether any of these bacterial isolates developed antibiotic resistances over time. Cefazolin, the most frequently used antibiotic in monkeys in this study, was the only antimicrobial out of 41 agents tested to which bacteria developed resistance over time. These results indicate that percutaneous implants are associated with a generalized inflammatory state, multiple bacterial species are present at the implant site, and these bacteria may contribute to the inflammatory response.

Measurement, manipulation and modeling of brain-wide neural population dynamics

Neural recording technologies increasingly enable simultaneous measurement of neural activity from multiple brain areas. To gain insight into distributed neural computations, a commensurate advance in experimental and analytical methods is necessary. We discuss two opportunities towards this end: the manipulation and modeling of neural population dynamics.

The temporal pattern of Intracortical Microstimulation pulses elicits distinct temporal and spatial recruitment of cortical neuropil and neurons

OBJECTIVE

The spacing or distribution of stimulation pulses of therapeutic neurostimulation waveforms-referred to here as the Temporal Pattern (TP)-has emerged as an important parameter for tuning the response to deep-brain stimulation and intracortical microstimulation (ICMS). While it has long been assumed that modulating the TP of ICMS may be effective by altering the rate coding of the neural response, it is unclear how it alters the neural response at the neural network level. The present study is designed to elucidate the neural response to TP at the network level.

APPROACH

We use in vivo two-photon imaging of ICMS in mice expressing the calcium sensor Thy1-GCaMP or the glutamate sensor hSyn-iGluSnFr to examine the layer II/III neural response to stimulations with different TPs. We study the neuronal calcium and glutamate response to TPs with the same average frequency (10Hz) and same total charge injection, but varying degrees of bursting. We also investigate one control pattern with an average frequency of 100Hz and 10X the charge injection.

MAIN RESULTS

Stimulation trains with the same average frequency (10 Hz) and same total charge injection but distinct temporal patterns recruits distinct sets of neurons. More-than-half (60% of 309 cells) prefer one temporal pattern over the other. Despite their distinct spatial recruitment patterns, both cells exhibit similar ability to follow 30s trains of both TPs without failing, and they exhibit similar levels of glutamate release during stimulation. Both neuronal calcium and glutamate release train to the bursting TP pattern (~21-fold increase in relative power at the frequency of bursting. Bursting also results in a statistically significant elevation in the correlation between somatic calcium activity and neuropil activity, which we explore as a metric for inhibitory-excitatory tone. Interestingly, soma-neuropil correlation during the bursting pattern is a statistically significant predictor of cell preference for TP, which exposes a key link between inhibitory-excitatory tone. Finally, using mesoscale imaging, we show that both TPs result in distal inhibition during stimulation, which reveals complex spatial and temporal interactions between temporal pattern and inhibitory-excitatory tone in ICMS.

SIGNIFICANCE

Our results may ultimately suggest that TP is a valuable parameter space to modulate inhibitory-excitatory tone as well as distinct network activity in ICMS. This presents a broader mechanism of action than rate coding, as previously thought. By implicating these additional mechanisms, TP may have broader utility in the clinic and should be pursued to expand the efficacy of ICMS therapies.

Recurrent circuit dynamics underlie persistent activity in the macaque frontoparietal network

During delayed oculomotor response tasks, neurons in the lateral intraparietal area (LIP) and the frontal eye fields (FEF) exhibit persistent activity that reflects the active maintenance of behaviorally relevant information. Despite many computational models of the mechanisms of persistent activity, there is a lack of circuit-level data from the primate to inform the theories. To fill this gap, we simultaneously recorded ensembles of neurons in both LIP and FEF while macaques performed a memory-guided saccade task. A population encoding model revealed strong and symmetric long-timescale recurrent excitation between LIP and FEF. Unexpectedly, LIP exhibited stronger local functional connectivity than FEF, and many neurons in LIP had longer network and intrinsic timescales. The differences in connectivity could be explained by the strength of recurrent dynamics in attractor networks. These findings reveal reciprocal multi-area circuit dynamics in the frontoparietal network during persistent activity and lay the groundwork for quantitative comparisons to theoretical models.

A dynamic, imperturbable link between midbrain activity and saccade velocity

We make a saccadic eye movement once every few hundred milliseconds; however, the neural control of saccade execution is not fully understood. Dynamic, moment-by-moment variations in saccade velocity are typically thought to be controlled by neurons in the lower, but not the upper regions of the brainstem. In a recent report, Smalianchuk et al. (Smalianchuk I, Jagadisan UK, Gandhi NJ. J Neurosci 38: 10156-10167, 2018) provided strong evidence for a role of the superior colliculus, a midbrain structure, in the instantaneous control of saccade velocity, suggesting the revision of long-standing models of oculomotor control.

Saccade metrics reflect decision-making dynamics during urgent choices

A perceptual judgment is typically characterized by constructing psychometric and chronometric functions, i.e., by mapping the accuracies and reaction times of motor choices as functions of a sensory stimulus feature dimension. Here, we show that various saccade metrics (e.g., peak velocity) are similarly modulated as functions of sensory cue viewing time during performance of an urgent-decision task. Each of the newly discovered functions reveals the dynamics of the perceptual evaluation process inherent to the underlying judgment. Remarkably, saccade peak velocity correlates with statistical decision confidence, suggesting that saccade kinematics reflect the degree of certainty with which an urgent perceptual decision is made. The data were explained by a race-to-threshold model that also replicates standard performance measures and cortical oculomotor neuronal activity in the task. The results indicate that, although largely stereotyped, saccade metrics carry subtle but reliable traces of the underlying cognitive processes that give rise to each oculomotor choice.

ERAASR: an algorithm for removing electrical stimulation artifacts from multielectrode array recordings

OBJECTIVE

Electrical stimulation is a widely used and effective tool in systems neuroscience, neural prosthetics, and clinical neurostimulation. However, electrical artifacts evoked by stimulation prevent the detection of spiking activity on nearby recording electrodes, which obscures the neural population response evoked by stimulation. We sought to develop a method to clean artifact-corrupted electrode signals recorded on multielectrode arrays in order to recover the underlying neural spiking activity.

APPROACH

We created an algorithm, which performs estimation and removal of array artifacts via sequential principal components regression (ERAASR). This approach leverages the similar structure of artifact transients, but not spiking activity, across simultaneously recorded channels on the array, across pulses within a train, and across trials. The ERAASR algorithm requires no special hardware, imposes no requirements on the shape of the artifact or the multielectrode array geometry, and comprises sequential application of straightforward linear methods with intuitive parameters. The approach should be readily applicable to most datasets where stimulation does not saturate the recording amplifier.

MAIN RESULTS

The effectiveness of the algorithm is demonstrated in macaque dorsal premotor cortex using acute linear multielectrode array recordings and single electrode stimulation. Large electrical artifacts appeared on all channels during stimulation. After application of ERAASR, the cleaned signals were quiescent on channels with no spontaneous spiking activity, whereas spontaneously active channels exhibited evoked spikes which closely resembled spontaneously occurring spiking waveforms.

SIGNIFICANCE

We hope that enabling simultaneous electrical stimulation and multielectrode array recording will help elucidate the causal links between neural activity and cognition and facilitate naturalistic sensory protheses.

Transient Pupil Dilation after Subsaccadic Microstimulation of Primate Frontal Eye Fields

Pupillometry provides a simple and noninvasive index for a variety of cognitive processes, including perception, attention, task consolidation, learning, and memory. The neural substrates by which such cognitive processes influence pupil diameter remain somewhat unclear, although cortical inputs to the locus coeruleus mediating arousal are likely involved. Changes in pupil diameter also accompany covert orienting; hence the oculomotor system may provide an alternative substrate for cognitive influences on pupil diameter. Here, we show that low-level electrical microstimulation of the primate frontal eye fields (FEFs), a cortical component of the oculomotor system strongly connected to the intermediate layers of the superior colliculus (SCi), evoked robust pupil dilation even in the absence of evoked saccades. The magnitude of such dilation scaled with increases in stimulation parameters, depending strongly on the intensity and number of pulses. Although there are multiple pathways by which FEF stimulation could cause pupil dilation, the timing and profile of dilation closely resembled that evoked by SCi stimulation. Moreover, pupil dilation evoked from the FEFs increased when presumed oculomotor activity was higher at the time of stimulation. Our findings implicate the oculomotor system as a potential substrate for how cognitive processes can influence pupil diameter. We suggest that a pathway from the frontal cortex through the SCi operates in parallel with frontal inputs to arousal circuits to regulate task-dependent modulation of pupil diameter, perhaps indicative of an organization wherein one pathway assumes primacy for a given cognitive process.

SIGNIFICANCE STATEMENT

Pupillometry (the measurement of pupil diameter) provides a simple and noninvasive index for a variety of cognitive processes, offering a biomarker that has value in both health and disease. But how do cognitive processes influence pupil diameter? Here, we show that low-level stimulation of the primate frontal eye fields can induce robust pupil dilation without saccades. Pupil dilation scaled with the number and intensity of stimulation pulses and varied with endogenous oculomotor activity at the time of stimulation. The oculomotor system therefore provides a plausible pathway by which cognitive processes may influence pupil diameter, perhaps operating in conjunction with systems regulating arousal.

The difference between electrical microstimulation and direct electrical stimulation – towards new opportunities for innovative functional brain mapping?

Both electrical microstimulation (EMS) and direct electrical stimulation (DES) of the brain are used to perform functional brain mapping. EMS is applied to animal fundamental neuroscience experiments, whereas DES is performed in the operating theatre on neurosurgery patients. The objective of the present review was to shed new light on electrical stimulation techniques in brain mapping by comparing EMS and DES. There is much controversy as to whether the use of DES during wide-awake surgery is the ‘gold standard’ for studying the brain function. As part of this debate, it is sometimes wrongly assumed that EMS and DES induce similar effects in the nervous tissues and have comparable behavioural consequences. In fact, the respective stimulation parameters in EMS and DES are clearly different. More surprisingly, there is no solid biophysical rationale for setting the stimulation parameters in EMS and DES; this may be due to historical, methodological and technical constraints that have limited the experimental protocols and prompted the use of empirical methods. In contrast, the gap between EMS and DES highlights the potential for new experimental paradigms in electrical stimulation for functional brain mapping. In view of this gap and recent technical developments in stimulator design, it may now be time to move towards alternative, innovative protocols based on the functional stimulation of peripheral nerves (for which a more solid theoretical grounding exists).

Behavioral assessment of sensitivity to intracortical microstimulation of primate somatosensory cortex

Intracortical microstimulation (ICMS) is a powerful tool to investigate the functional role of neural circuits and may provide a means to restore sensation for patients for whom peripheral stimulation is not an option. In a series of psychophysical experiments with nonhuman primates, we investigate how stimulation parameters affect behavioral sensitivity to ICMS. Specifically, we deliver ICMS to primary somatosensory cortex through chronically implanted electrode arrays across a wide range of stimulation regimes. First, we investigate how the detectability of ICMS depends on stimulation parameters, including pulse width, frequency, amplitude, and pulse train duration. Then, we characterize the degree to which ICMS pulse trains that differ in amplitude lead to discriminable percepts across the range of perceptible and safe amplitudes. We also investigate how discriminability of pulse amplitude is modulated by other stimulation parameters-namely, frequency and duration. Perceptual judgments obtained across these various conditions will inform the design of stimulation regimes for neuroscience and neuroengineering applications.

Temporally-patterned deep brain stimulation in a mouse model of multiple traumatic brain injury

We report that mice with closed-head multiple traumatic brain injury (TBI) show a decrease in the motoric aspects of generalized arousal, as measured by automated, quantitative behavioral assays. Further, we found that temporally-patterned deep brain stimulation (DBS) can increase generalized arousal and spontaneous motor activity in this mouse model of TBI. This arousal increase is input-pattern-dependent, as changing the temporal pattern of DBS can modulate its effect on motor activity. Finally, an extensive examination of mouse behavioral capacities, looking for deficits in this model of TBI, suggest that the strongest effects of TBI in this model are found in the initiation of any kind of movement.

Proposing a two-level stochastic model for epileptic seizure genesis

By assuming the brain as a multi-stable system, different scenarios have been introduced for transition from normal to epileptic state. But, the path through which this transition occurs is under debate. In this paper a stochastic model for seizure genesis is presented that is consistent with all scenarios: a two-level spontaneous seizure generation model is proposed in which, in its first level the behavior of physiological parameters is modeled with a stochastic process. The focus is on some physiological parameters that are essential in simulating different activities of ElectroEncephaloGram (EEG), i.e., excitatory and inhibitory synaptic gains of neuronal populations. There are many depth-EEG models in which excitatory and inhibitory synaptic gains are the adjustable parameters. Using one of these models at the second level, our proposed seizure generator is complete. The suggested stochastic model of first level is a hidden Markov process whose transition matrices are obtained through analyzing the real parameter sequences of a seizure onset area. These real parameter sequences are estimated from real depth-EEG signals via applying a parameter identification algorithm. In this paper both short-term and long-term validations of the proposed model are done. The long-term synthetic depth-EEG signals simulated by this model can be taken as a suitable tool for comparing different seizure prediction algorithms.

Tracking the eye non-invasively: simultaneous comparison of the scleral search coil and optical tracking techniques in the macaque monkey

From human perception to primate neurophysiology, monitoring eye position is critical to the study of vision, attention, oculomotor control, and behavior. Two principal techniques for the precise measurement of eye position—the long-standing sclera-embedded search coil and more recent optical tracking techniques—are in use in various laboratories, but no published study compares the performance of the two methods simultaneously in the same primates. Here we compare two popular systems—a sclera-embedded search coil from C-N-C Engineering and the EyeLink 1000 optical system from SR Research—by recording simultaneously from the same eye in the macaque monkey while the animal performed a simple oculomotor task. We found broad agreement between the two systems, particularly in positional accuracy during fixation, measurement of saccade amplitude, detection of fixational saccades, and sensitivity to subtle changes in eye position from trial to trial. Nonetheless, certain discrepancies persist, particularly elevated saccade peak velocities, post-saccadic ringing, influence of luminance change on reported position, and greater sample-to-sample variation in the optical system. Our study shows that optical performance now rivals that of the search coil, rendering optical systems appropriate for many if not most applications. This finding is consequential, especially for animal subjects, because the optical systems do not require invasive surgery for implantation and repair of search coils around the eye. Our data also allow laboratories using the optical system in human subjects to assess the strengths and limitations of the technique for their own applications.

The relative impact of microstimulation parameters on movement generation

Microstimulation is widely used in neurophysiology to characterize brain areas with behavior and in clinical therapeutics to treat neurological disorder. Current intensity and frequency, which respectively influence activation patterns in spatial and temporal domains, are typically selected to elicit a desired response, but their effective influence on behavior has not been thoroughly examined. We delivered microstimulation to the primate superior colliculus while systematically varying each parameter to capture effects of a large range of parameter space. We found that frequency was more effective in driving output properties, whereas properties changed gradually with intensity. Interestingly, when different parameter combinations were matched for total charge, effects on behavioral properties became seemingly equivalent. This study provides a first level resource for choosing desired parameter ranges to effectively manipulate behavior. It also provides insights into interchangeability of parameters, which can assist clinical microstimulation that looks to appropriately control behavior within designated constraints, such as power consumption.

Insights into cortical mechanisms of behavior from microstimulation experiments

Even the simplest behaviors depend on a large number of neurons that are distributed across many brain regions. Because electrical microstimulation can change the activity of localized subsets of neurons, it has provided valuable evidence that specific neurons contribute to particular behaviors. Here we review what has been learned about cortical function from behavioral studies using microstimulation in animals and humans. Experiments that examine how microstimulation affects the perception of stimuli have shown that the effects of microstimulation are usually highly specific and can be related to the stimuli preferred by neurons at the stimulated site. Experiments that ask subjects to detect cortical microstimulation in the absence of other stimuli have provided further insights. Although subjects typically can detect microstimulation of primary sensory or motor cortex, they are generally unable to detect stimulation of most of cortex without extensive practice. With practice, however, stimulation of any part of cortex can become detected. These training effects suggest that some patterns of cortical activity cannot be readily accessed to guide behavior, but that the adult brain retains enough plasticity to learn to process novel patterns of neuronal activity arising anywhere in cortex.

An electric field model for prediction of somatosensory (S1) cortical field potentials induced by ventral posterior lateral (VPL) thalamic microstimulation

Microstimulation (MiSt) is used experimentally and clinically to activate localized populations of neural elements. However, it is difficult to predict-and subsequently control-neural responses to simultaneous current injection through multiple electrodes in an array. This is due to the unknown locations of neuronal elements in the extracellular medium that are excited by the superposition of multiple parallel current sources. We, therefore, propose a model that maps the computed electric field in the 3-D space surrounding the stimulating electrodes in one brain region to the local field potential (LFP) fluctuations evoked in a downstream region. Our model is trained with the recorded LFP waveforms in the primary somatosensory cortex (S1) resulting from MiSt applied in multiple electrode configurations in the ventral posterolateral nucleus (VPL) of the quiet awake rat. We then predict the cortical responses to MiSt in "novel" electrode configurations, a result that suggests that this technique could aid in the design of spatially optimized MiSt patterns through a multielectrode array.

Probing neural circuitry and function with electrical microstimulation

Since the discovery of the nervous system's electrical excitability more than 200 years ago, neuroscientists have used electrical stimulation to manipulate brain activity in order to study its function. Microstimulation has been a valuable technique for probing neural circuitry and identifying networks of neurons that underlie perception, movement and cognition. In this review, we focus on the use of stimulation in behaving primates, an experimental system that permits causal inferences to be made about the effect of stimulation-induced activity on the resulting behaviour or neural signals elsewhere in the brain.

A Brief Survey of Computational Models of Normal and Epileptic EEG Signals: A Guideline to Model-based Seizure Prediction

In recent decades, seizure prediction has caused a lot of research in both signal processing and the neuroscience field. The researches have tried to enhance the conventional seizure prediction algorithms such that the rate of the false alarms be appropriately small, so that seizures can be predicted according to clinical standards. To date, none of the proposed algorithms have been sufficiently adequate. In this article we show that in considering the mechanism of the generation of seizures, the prediction results may be improved. For this purpose, an algorithm based on the identification of the parameters of a physiological model of seizures is introduced. Some models of electroencephalographic (EEG) signals that can also be potentially considered as models of seizure and some developed seizure models are reviewed. As an example the model of depth-EEG signals, proposed by Wendling, is studied and is shown to be a suitable model.

Temporal patterning of pulses during deep brain stimulation affects central nervous system arousal

Regulation of CNS arousal is important for a wide variety of functions, including the initiation of all motivated behaviors. Usually studied with pharmacological or hormonal tools, CNS arousal can also be elevated by deep brain stimulation (DBS), in the human brain and in animals. The effectiveness of DBS is conventionally held to depend on pulse width, frequency, amplitude and stimulation duration. We demonstrate a novel approach for testing the effectiveness of DBS to increase arousal in intact female mice: all of the foregoing parameters are held constant. Only the temporal patterning of the pulses within the stimulation is varied. To create differentially patterned pulse trains, a deterministic nonlinear dynamic equation was used to generate a series of pulses with a predetermined average frequency. Three temporal patterns of stimulation were defined: two nonlinear patterns, Nonlinear1 (NL1) and Nonlinear2 (NL2), and the conventional pattern, Fixed Frequency (FF). Female mice with bilateral monopolar electrodes were observed before, during and after hippocampal or medial thalamic stimulation. NL1 hippocampal stimulation was significantly more effective at increasing behavioral arousal than either FF or NL2; however, FF and NL2 stimulation of the medial thalamus were more effective than NL1. During the same experiments, we recorded an unpredicted increase in the spectral power of slow waves in the cortical EEG. Our data comprise the first demonstration that the temporal pattern of DBS can be used to elevate its effectiveness, and also point the way toward the use of nonlinear dynamics in the exploration of means to optimize DBS.

Microstimulation reveals limits in detecting different signals from a local cortical region

Behavioral performance depends on the activity of neurons in sensory cortex, but little is known about the brain's capacity to access specific neuronal signals to guide behavior. Even the individual sensory neurons that are most sensitive to a relevant stimulus are only weakly correlated with behavior, suggesting that behavioral decisions are based on the combined activity of groups of neurons with sensitivities well matched to task demands. To explore how flexibly different patterns of activity can be accessed from a given cortical region, we trained animals to detect electrical microstimulation of local V1 sites. By allowing the animals to become expert at the detection of microstimulation of specific V1 sites that corresponded to particular retinotopic locations, we could measure the effects of that training on the ability of those sites to support the detection of visual stimuli. Training to detect electrical activation caused a large, reversible, retinotopically localized impairment of thresholds for detecting visual stimuli. Retraining on visual detection restored normal thresholds and in turn impaired thresholds for detecting microstimulation. These results suggest that there are substantial limits to the types of signals for which a local cortical region can be simultaneously optimized.

Direct activation of sparse, distributed populations of cortical neurons by electrical microstimulation

For over a century, electrical microstimulation has been the most direct method for causally linking brain function with behavior. Despite this long history, it is still unclear how the activity of neural populations is affected by stimulation. For example, there is still no consensus on where activated cells lie or on the extent to which neural processes such as passing axons near the electrode are also activated. Past studies of this question have proven difficult because microstimulation interferes with electrophysiological recordings, which in any case provide only coarse information about the location of activated cells. We used two-photon calcium imaging, an optical method, to circumvent these hurdles. We found that microstimulation sparsely activates neurons around the electrode, sometimes as far as millimeters away, even at low currents. Our results indicate that the pattern of activated neurons likely arises from the direct activation of axons in a volume tens of microns in diameter.