Comparison of brain–computer interface decoding algorithms in open-loop and closed-loop control

Neuroprosthetic devices such as a computer cursor can be controlled by the activity of cortical neurons when an appropriate algorithm is used to decode motor intention. Algorithms which have been proposed for this purpose range from the simple population vector algorithm (PVA) and optimal linear estimator (OLE) to various versions of Bayesian decoders. Although Bayesian decoders typically provide the most accurate off-line reconstructions, it is not known which model assumptions in these algorithms are critical for improving decoding performance. Furthermore, it is not necessarily true that improvements (or deficits) in off-line reconstruction will translate into improvements (or deficits) in on-line control, as the subject might compensate for the specifics of the decoder in use at the time. Here we show that by comparing the performance of nine decoders, assumptions about uniformly distributed preferred directions and the way the cursor trajectories are smoothed have the most impact on decoder performance in off-line reconstruction, while assumptions about tuning curve linearity and spike count variance play relatively minor roles. In on-line control, subjects compensate for directional biases caused by non-uniformly distributed preferred directions, leaving cursor smoothing differences as the largest single algorithmic difference driving decoder performance.     

A symbiotic brain-machine interface through value-based decision making

Background

In the development of Brain Machine Interfaces (BMIs), there is a great need to enable users to interact with changing environments during the activities of daily life. It is expected that the number and scope of the learning tasks encountered during interaction with the environment as well as the pattern of brain activity will vary over time. These conditions, in addition to neural reorganization, pose a challenge to decoding neural commands for BMIs. We have developed a new BMI framework in which a computational agent symbiotically decoded users'' intended actions by utilizing both motor commands and goal information directly from the brain through a continuous Perception-Action-Reward Cycle (PARC).

Methodology

The control architecture designed was based on Actor-Critic learning, which is a PARC-based reinforcement learning method. Our neurophysiology studies in rat models suggested that Nucleus Accumbens (NAcc) contained a rich representation of goal information in terms of predicting the probability of earning reward and it could be translated into an evaluative feedback for adaptation of the decoder with high precision. Simulated neural control experiments showed that the system was able to maintain high performance in decoding neural motor commands during novel tasks or in the presence of reorganization in the neural input. We then implanted a dual micro-wire array in the primary motor cortex (M1) and the NAcc of rat brain and implemented a full closed-loop system in which robot actions were decoded from the single unit activity in M1 based on an evaluative feedback that was estimated from NAcc.

Conclusions

Our results suggest that adapting the BMI decoder with an evaluative feedback that is directly extracted from the brain is a possible solution to the problem of operating BMIs in changing environments with dynamic neural signals. During closed-loop control, the agent was able to solve a reaching task by capturing the action and reward interdependency in the brain.     

Spatiotemporal Spike Coding of Behavioral Adaptation in the Dorsal Anterior Cingulate Cortex

The frontal cortex controls behavioral adaptation in environments governed by complex rules. Many studies have established the relevance of firing rate modulation after informative events signaling whether and how to update the behavioral policy. However, whether the spatiotemporal features of these neuronal activities contribute to encoding imminent behavioral updates remains unclear. We investigated this issue in the dorsal anterior cingulate cortex (dACC) of monkeys while they adapted their behavior based on their memory of feedback from past choices. We analyzed spike trains of both single units and pairs of simultaneously recorded neurons using an algorithm that emulates different biologically plausible decoding circuits. This method permits the assessment of the performance of both spike-count and spike-timing sensitive decoders. In response to the feedback, single neurons emitted stereotypical spike trains whose temporal structure identified informative events with higher accuracy than mere spike count. The optimal decoding time scale was in the range of 70–200 ms, which is significantly shorter than the memory time scale required by the behavioral task. Importantly, the temporal spiking patterns of single units were predictive of the monkeys’ behavioral response time. Furthermore, some features of these spiking patterns often varied between jointly recorded neurons. All together, our results suggest that dACC drives behavioral adaptation through complex spatiotemporal spike coding. They also indicate that downstream networks, which decode dACC feedback signals, are unlikely to act as mere neural integrators.     

A Real-Time Brain-Machine Interface Combining Motor Target and Trajectory Intent Using an Optimal Feedback Control Design

Real-time brain-machine interfaces (BMI) have focused on either estimating the continuous movement trajectory or target intent. However, natural movement often incorporates both. Additionally, BMIs can be modeled as a feedback control system in which the subject modulates the neural activity to move the prosthetic device towards a desired target while receiving real-time sensory feedback of the state of the movement. We develop a novel real-time BMI using an optimal feedback control design that jointly estimates the movement target and trajectory of monkeys in two stages. First, the target is decoded from neural spiking activity before movement initiation. Second, the trajectory is decoded by combining the decoded target with the peri-movement spiking activity using an optimal feedback control design. This design exploits a recursive Bayesian decoder that uses an optimal feedback control model of the sensorimotor system to take into account the intended target location and the sensory feedback in its trajectory estimation from spiking activity. The real-time BMI processes the spiking activity directly using point process modeling. We implement the BMI in experiments consisting of an instructed-delay center-out task in which monkeys are presented with a target location on the screen during a delay period and then have to move a cursor to it without touching the incorrect targets. We show that the two-stage BMI performs more accurately than either stage alone. Correct target prediction can compensate for inaccurate trajectory estimation and vice versa. The optimal feedback control design also results in trajectories that are smoother and have lower estimation error. The two-stage decoder also performs better than linear regression approaches in offline cross-validation analyses. Our results demonstrate the advantage of a BMI design that jointly estimates the target and trajectory of movement and more closely mimics the sensorimotor control system.     

Spike-Threshold Adaptation Predicted by Membrane Potential Dynamics In Vivo

Neurons encode information in sequences of spikes, which are triggered when their membrane potential crosses a threshold. In vivo, the spiking threshold displays large variability suggesting that threshold dynamics have a profound influence on how the combined input of a neuron is encoded in the spiking. Threshold variability could be explained by adaptation to the membrane potential. However, it could also be the case that most threshold variability reflects noise and processes other than threshold adaptation. Here, we investigated threshold variation in auditory neurons responses recorded in vivo in barn owls. We found that spike threshold is quantitatively predicted by a model in which the threshold adapts, tracking the membrane potential at a short timescale. As a result, in these neurons, slow voltage fluctuations do not contribute to spiking because they are filtered by threshold adaptation. More importantly, these neurons can only respond to input spikes arriving together on a millisecond timescale. These results demonstrate that fast adaptation to the membrane potential captures spike threshold variability in vivo.     

A computational framework for cortical learning

Recent physiological findings have revealed that long-term adaptation of the synaptic strengths between cortical pyramidal neurons depends on the temporal order of presynaptic and postsynaptic spikes, which is called spike-timing-dependent plasticity (STDP) or temporally asymmetric Hebbian (TAH) learning. Here I prove by analytical means that a physiologically plausible variant of STDP adapts synaptic strengths such that the presynaptic spikes predict the postsynaptic spikes with minimal error. This prediction error model of STDP implies a mechanism for cortical memory: cortical tissue learns temporal spike patterns if these spike patterns are repeatedly elicited in a set of pyramidal neurons. The trained network finishes these patterns if their beginnings are presented, thereby recalling the memory. Implementations of the proposed algorithms may be useful for applications in voice recognition and computer vision.     

多通道神经元锋电位检测和分类的新方法

大脑神经元胞外单细胞动作电位(即锋电位)的检测和分类是提取神经元脉冲序列、研究神经系统信息处理机制的关键．为了提高锋电位的检出率和分类的正确性,设计了一种处理多通道锋电位记录信号的算法,用于分析微电极阵列记录的大鼠海马神经元锋电位信号,电极阵列上的测量点排列紧密,4个通道可以同时记录到来自相同神经元的信号．该算法首先利用一种多通道阈值检测法检出四通道记录信号中的锋电位,然后利用一种基于复合锋电位的主成分特征参数分类法将锋电位分类．仿真数据和实验记录信号的检验结果表明：与相应的单通道算法相比,该算法的锋电位检出率和分类的正确性显著提高,并且可以增加单次实验测得的神经元数目．因此,该算法为实现神经元锋电位的自动检测提供了一种简单有效的新 方法．     

Mathematical analysis of the possibility of detecting correlation between spike trains of weakly interacting neurons

The number of spikes which must be recorded in order to detect significant correlation between spike trains of two synaptically connected neurons was estimated by a mathematical model. Dependence of this number of spikes on importance of interneuronal connection (measured as the amplitude of the EPSP evoked by a single spike of the input neuron in the output cell) and on the intensity of total spontaneous excitatory influences on the output neuron and on its own parameters was studied. For cells which corresponded in the weight of connections between them, their intrinsic parameters, and characteristics of spontaneous activity to real spinal neurons, the necessary number of spikes was 10⁷–10⁸. An increase in amplitude of the single EPSP and also a decrease in the intensity of the input spontaneous spike train and parameters of after-hyperpolarization of the postsynaptic neuron led to a decrease in the number of spikes necessary for the detection of significant correlation. On the basis of the results of this and previous investigations the possible principles for construction of a spinal locomotor generator are discussed.A. A. Bogomolets Institute of Physiology, Academy of Sciences of the Ukrainian SSR, Kiev. Translated from Neirofiziologiya, Vol. 12, No. 3, pp. 290–296, May–June, 1980.     

Encoder-Decoder Optimization for Brain-Computer Interfaces

Neuroprosthetic brain-computer interfaces are systems that decode neural activity into useful control signals for effectors, such as a cursor on a computer screen. It has long been recognized that both the user and decoding system can adapt to increase the accuracy of the end effector. Co-adaptation is the process whereby a user learns to control the system in conjunction with the decoder adapting to learn the user''s neural patterns. We provide a mathematical framework for co-adaptation and relate co-adaptation to the joint optimization of the user''s control scheme ("encoding model") and the decoding algorithm''s parameters. When the assumptions of that framework are respected, co-adaptation cannot yield better performance than that obtainable by an optimal initial choice of fixed decoder, coupled with optimal user learning. For a specific case, we provide numerical methods to obtain such an optimized decoder. We demonstrate our approach in a model brain-computer interface system using an online prosthesis simulator, a simple human-in-the-loop pyschophysics setup which provides a non-invasive simulation of the BCI setting. These experiments support two claims: that users can learn encoders matched to fixed, optimal decoders and that, once learned, our approach yields expected performance advantages.     

Inheritance of five quantitative characters of bread wheat

Summary Heritability estimates of five characters of the wheat plant were studied in five crosses involving six cultivars of bread wheat. Parents, F₁, F₂ and backcrosses to both parents were used in the estimation of the genetic parameters.Heritability was low for number of fertile spikes/plant, moderate for number of spikelets/spike, number of kernels/spike, 1000-kernel weight and moderately high for number of kernels/spike. Evidence for mainly nonadditive gene effects were observed in the expression of number of fertile spikes and 1000-kernel weight. Although nonadditives contributed to a lesser degree to the gene action, additives seemed to be the most important genetic expression regulating number of spikelets/spike, number of kernels/spike, and number of kernels/spikelet. Except for number of fertile spikes/plant, selection in F₂-populations seems to be promising.     

Correlations in state space can cause sub-optimal adaptation of optimal feedback control models

Control of our movements is apparently facilitated by an adaptive internal model in the cerebellum. It was long thought that this internal model implemented an adaptive inverse model and generated motor commands, but recently many reject that idea in favor of a forward model hypothesis. In theory, the forward model predicts upcoming state during reaching movements so the motor cortex can generate appropriate motor commands. Recent computational models of this process rely on the optimal feedback control (OFC) framework of control theory. OFC is a powerful tool for describing motor control, it does not describe adaptation. Some assume that adaptation of the forward model alone could explain motor adaptation, but this is widely understood to be overly simplistic. However, an adaptive optimal controller is difficult to implement. A reasonable alternative is to allow forward model adaptation to ‘re-tune’ the controller. Our simulations show that, as expected, forward model adaptation alone does not produce optimal trajectories during reaching movements perturbed by force fields. However, they also show that re-optimizing the controller from the forward model can be sub-optimal. This is because, in a system with state correlations or redundancies, accurate prediction requires different information than optimal control. We find that adding noise to the movements that matches noise found in human data is enough to overcome this problem. However, since the state space for control of real movements is far more complex than in our simple simulations, the effects of correlations on re-adaptation of the controller from the forward model cannot be overlooked.     

Slow and Spike Potentials Recorded from Retinula Cells of the Honeybee Drone in Response to Light

Responses to light recorded by means of intracellular microelectrodes in isolated heads kept in oxygenated Ringer solution consist of a slow depolarization. Light adaptation increases the rates of depolarization and repolarization and decreases the amplitude of the response. Qualitatively these changes are similar to those observed in Limulus by Fuortes and Hodgkin. They are rapidly reversible during dark adaptation. In retinula cells of the drone eye a large single spike is recorded superimposed on the rising phase of the slow potential. The spike is a regenerative phenomenon; it can be triggered with electric current and is markedly reduced, sometimes abolished by tetrodotoxin. In rare cases cells were found which responded to light with a train of spikes. This behavior was only found under "unusual" experimental conditions; i.e., towards the end of a long experiment, during impalement, or at the beginning of responses to steps of strongly light-adapted preparations.     

Reliability of spike and burst firing in thalamocortical relay cells

The reliability and precision of the timing of spikes in a spike train is an important aspect of neuronal coding. We investigated reliability in thalamocortical relay (TCR) cells in the acute slice and also in a Morris-Lecar model with several extensions. A frozen Gaussian noise current, superimposed on a DC current, was injected into the TCR cell soma. The neuron responded with spike trains that showed trial-to-trial variability, due to amongst others slow changes in its internal state and the experimental setup. The DC current allowed to bring the neuron in different states, characterized by a well defined membrane voltage (between ?80 and ?50 mV) and by a specific firing regime that on depolarization gradually shifted from a predominantly bursting regime to a tonic spiking regime. The filtered frozen white noise generated a spike pattern output with a broad spike interval distribution. The coincidence factor and the Hunter and Milton measure were used as reliability measures of the output spike train. In the experimental TCR cell as well as the Morris-Lecar model cell the reliability depends on the shape (steepness) of the current input versus spike frequency output curve. The model also allowed to study the contribution of three relevant ionic membrane currents to reliability: a T-type calcium current, a cation selective h-current and a calcium dependent potassium current in order to allow bursting, investigate the consequences of a more complex current-frequency relation and produce realistic firing rates. The reliability of the output of the TCR cell increases with depolarization. In hyperpolarized states bursts are more reliable than single spikes. The analytically derived relations were capable to predict several of the experimentally recorded spike features.     

Interspike interval correlations in neuron models with adaptation and correlated noise

The generation of neural action potentials (spikes) is random but nevertheless may result in a rich statistical structure of the spike sequence. In particular, contrary to the popular renewal assumption of theoreticians, the intervals between adjacent spikes are often correlated. Experimentally, different patterns of interspike-interval correlations have been observed and computational studies have identified spike-frequency adaptation and correlated noise as the two main mechanisms that can lead to such correlations. Analytical studies have focused on the single cases of either correlated (colored) noise or adaptation currents in combination with uncorrelated (white) noise. For low-pass filtered noise or adaptation, the serial correlation coefficient can be approximated as a single geometric sequence of the lag between the intervals, providing an explanation for some of the experimentally observed patterns. Here we address the problem of interval correlations for a widely used class of models, multidimensional integrate-and-fire neurons subject to a combination of colored and white noise sources and a spike-triggered adaptation current. Assuming weak noise, we derive a simple formula for the serial correlation coefficient, a sum of two geometric sequences, which accounts for a large class of correlation patterns. The theory is confirmed by means of numerical simulations in a number of special cases including the leaky, quadratic, and generalized integrate-and-fire models with colored noise and spike-frequency adaptation. Furthermore we study the case in which the adaptation current and the colored noise share the same time scale, corresponding to a slow stochastic population of adaptation channels; we demonstrate that our theory can account for a nonmonotonic dependence of the correlation coefficient on the channel’s time scale. Another application of the theory is a neuron driven by network-noise-like fluctuations (green noise). We also discuss the range of validity of our weak-noise theory and show that by changing the relative strength of white and colored noise sources, we can change the sign of the correlation coefficient. Finally, we apply our theory to a conductance-based model which demonstrates its broad applicability.     

Nonlinear effects of intrinsic dynamics on temporal encoding in a model of avian auditory cortex

Neurons exhibit diverse intrinsic dynamics, which govern how they integrate synaptic inputs to produce spikes. Intrinsic dynamics are often plastic during development and learning, but the effects of these changes on stimulus encoding properties are not well known. To examine this relationship, we simulated auditory responses to zebra finch song using a linear-dynamical cascade model, which combines a linear spectrotemporal receptive field with a dynamical, conductance-based neuron model, then used generalized linear models to estimate encoding properties from the resulting spike trains. We focused on the effects of a low-threshold potassium current (K_LT) that is present in a subset of cells in the zebra finch caudal mesopallium and is affected by early auditory experience. We found that K_LT affects both spike adaptation and the temporal filtering properties of the receptive field. The direction of the effects depended on the temporal modulation tuning of the linear (input) stage of the cascade model, indicating a strongly nonlinear relationship. These results suggest that small changes in intrinsic dynamics in tandem with differences in synaptic connectivity can have dramatic effects on the tuning of auditory neurons.     

Real-Time Decision Fusion for Multimodal Neural Prosthetic Devices

Background

The field of neural prosthetics aims to develop prosthetic limbs with a brain-computer interface (BCI) through which neural activity is decoded into movements. A natural extension of current research is the incorporation of neural activity from multiple modalities to more accurately estimate the user''s intent. The challenge remains how to appropriately combine this information in real-time for a neural prosthetic device.

Methodology/Principal Findings

Here we propose a framework based on decision fusion, i.e., fusing predictions from several single-modality decoders to produce a more accurate device state estimate. We examine two algorithms for continuous variable decision fusion: the Kalman filter and artificial neural networks (ANNs). Using simulated cortical neural spike signals, we implemented several successful individual neural decoding algorithms, and tested the capabilities of each fusion method in the context of decoding 2-dimensional endpoint trajectories of a neural prosthetic arm. Extensively testing these methods on random trajectories, we find that on average both the Kalman filter and ANNs successfully fuse the individual decoder estimates to produce more accurate predictions.

Conclusions

Our results reveal that a fusion-based approach has the potential to improve prediction accuracy over individual decoders of varying quality, and we hope that this work will encourage multimodal neural prosthetics experiments in the future.     

Predicting spike timing of neocortical pyramidal neurons by simple threshold models

Neurons generate spikes reliably with millisecond precision if driven by a fluctuating current—is it then possible to predict the spike timing knowing the input? We determined parameters of an adapting threshold model using data recorded in vitro from 24 layer 5 pyramidal neurons from rat somatosensory cortex, stimulated intracellularly by a fluctuating current simulating synaptic bombardment in vivo. The model generates output spikes whenever the membrane voltage (a filtered version of the input current) reaches a dynamic threshold. We find that for input currents with large fluctuation amplitude, up to 75% of the spike times can be predicted with a precision of ±2 ms. Some of the intrinsic neuronal unreliability can be accounted for by a noisy threshold mechanism. Our results suggest that, under random current injection into the soma, (i) neuronal behavior in the subthreshold regime can be well approximated by a simple linear filter; and (ii) most of the nonlinearities are captured by a simple threshold process.     

A role for the cerebellum in the control of limb movement velocity

Accepting, rejecting or modifying the many different theories of the cerebellum's role in the control of movement requires an understanding of the signals encoded in the discharge of cerebellar neurons and how those signals are transformed by the cerebellar circuitry. Particularly challenging is understanding the sensory and motor signals carried by the two types of action potentials generated by cerebellar Purkinje cells, the simple spikes and complex spikes. Advances have been made in understanding this signal processing in the context of voluntary arm movements. Recent evidence suggests that mossy fiber afferents to the cerebellar cortex are a source of kinematic signals, providing information about movement direction and speed. In turn, the simple spike discharge of Purkinje cells integrates this mossy fiber information to generate a movement velocity signal. Complex spikes may signal errors in movement velocity. It is proposed that the cerebellum uses the signals carried by the simple and complex spike discharges to control movement velocity for both step and tracking arm movements.