Kelvin So

Subject-specific modulation of local field potential spectral power during brain–machine interface control in primates

OBJECTIVE Intracortical brain-machine interfaces (BMIs) have predominantly utilized spike activity as the control signal. However, an increasing number of studies have shown the utility of local field potentials (LFPs) for decoding motor related signals. Currently, it is unclear how well different LFP frequencies can serve as features for continuous, closed-loop BMI control. APPROACH We demonstrate 2D continuous LFP-based BMI control using closed-loop decoder adaptation, which adapts decoder parameters to subject-specific LFP feature modulations during BMI control. We trained two macaque monkeys to control a 2D cursor in a center-out task by modulating LFP power in the 0-150 Hz range. MAIN RESULTS While both monkeys attained control, they used different strategies involving different frequency bands. One monkey primarily utilized the low-frequency spectrum (0-80 Hz), which was highly correlated between channels, and obtained proficient performance even with a single channel. In contrast, the other monkey relied more on higher frequencies (80-150 Hz), which were less correlated between channels, and had greater difficulty with control as the number of channels decreased. We then restricted the monkeys to use only various sub-ranges (0-40, 40-80, and 80-150 Hz) of the 0-150 Hz band. Interestingly, although both monkeys performed better with some sub-ranges than others, they were able to achieve BMI control with all sub-ranges after decoder adaptation, demonstrating broad flexibility in the frequencies that could potentially be used for LFP-based BMI control. SIGNIFICANCE Overall, our results demonstrate proficient, continuous BMI control using LFPs and provide insight into the subject-specific spectral patterns of LFP activity modulated during control.

Assessing functional connectivity of neural ensembles using directed information

Neurons in the brain form highly complex networks through synaptic connections. Traditionally, functional connectivity between neurons has been explored using methods such as correlations, which do not contain any notion of directionality. Recently, an information-theoretic approach based on directed information theory has been proposed as a way to infer the direction of influence. However, it is still unclear whether this new approach provides any additional insight beyond conventional correlation analyses. In this paper, we present a modified procedure for estimating directed information and provide a comparison of results obtained using correlation analyses on both simulated and experimental data. Using physiologically realistic simulations, we demonstrate that directed information can outperform correlation in determining connections between neural spike trains while also providing directionality of the relationship, which cannot be assessed using correlation. Secondly, applying our method to rodent and primate data sets, we demonstrate that directed information can accurately estimate the conduction delay in connections between different brain structures. Moreover, directed information reveals connectivity structures that are not captured by correlations. Hence, directed information provides accurate and novel insights into the functional connectivity of neural ensembles that are applicable to data from neurophysiological studies in awake behaving animals.

Continuous closed-loop decoder adaptation with a recursive maximum likelihood algorithm allows for rapid performance acquisition in brain-machine interfaces

Closed-loop decoder adaptation (CLDA) is an emerging paradigm for both improving and maintaining online performance in brain-machine interfaces (BMIs). The time required for initial decoder training and any subsequent decoder recalibrations could be potentially reduced by performing continuous adaptation, in which decoder parameters are updated at every time step during these procedures, rather than waiting to update the decoder at periodic intervals in a more batch-based process. Here, we present recursive maximum likelihood (RML), a CLDA algorithm that performs continuous adaptation of a Kalman filter decoders parameters. We demonstrate that RML possesses a variety of useful properties and practical algorithmic advantages. First, we show how RML leverages the accuracy of updates based on a batch of data while still adapting parameters on every time step. Second, we illustrate how the RML algorithm is parameterized by a single, intuitive half-life parameter that can be used to adjust the rate of adaptation in real time. Third, we show how even when the number of neural features is very large, RMLs memory-efficient recursive update rules can be reformulated to also be computationally fast so that continuous adaptation is still feasible. To test the algorithm in closed-loop experiments, we trained three macaque monkeys to perform a center-out reaching task by using either spiking activity or local field potentials to control a 2D computer cursor. RML achieved higher levels of performance more rapidly in comparison to a previous CLDA algorithm that adapts parameters on a more intermediate timescale. Overall, our results indicate that RML is an effective CLDA algorithm for achieving rapid performance acquisition using continuous adaptation.

Redundant information encoding in primary motor cortex during natural and prosthetic motor control

Redundant encoding of information facilitates reliable distributed information processing. To explore this hypothesis in the motor system, we applied concepts from information theory to quantify the redundancy of movement-related information encoded in the macaque primary motor cortex (M1) during natural and neuroprosthetic control. Two macaque monkeys were trained to perform a delay center-out reaching task controlling a computer cursor under natural arm movement (manual control, ‘MC’), and using a brain-machine interface (BMI) via volitional control of neural ensemble activity (brain control, ‘BC’). During MC, we found neurons in contralateral M1 to contain higher and more redundant information about target direction than ipsilateral M1 neurons, consistent with the laterality of movement control. During BC, we found that the M1 neurons directly incorporated into the BMI (‘direct’ neurons) contained the highest and most redundant target information compared to neurons that were not incorporated into the BMI (‘indirect’ neurons). This effect was even more significant when comparing to M1 neurons of the opposite hemisphere. Interestingly, when we retrained the BMI to use ipsilateral M1 activity, we found that these neurons were more redundant and contained higher information than contralateral M1 neurons, even though ensembles from this hemisphere were previously less redundant during natural arm movement. These results indicate that ensembles most associated to movement contain highest redundancy and information encoding, which suggests a role for redundancy in proficient natural and prosthetic motor control.

Brain-machine interface control using broadband spectral power from local field potentials

Recent progress in brain-machine interfaces (BMIs) has shown tremendous improvements in task complexity and degree of control. In particular, closed-loop decoder adaptation (CLDA) has emerged as an effective paradigm for both improving and maintaining the performance of BMI systems. Here, we demonstrate the first reported use of a CLDA algorithm to rapidly achieve high-performance control of a BMI based on local field potentials (LFPs). We trained a non-human primate to control a 2-D computer cursor by modulating LFP activity to perform a center-out reaching task, while applying CLDA to adaptively update the decoder. We show that the subject is quickly able to readily reach and hold at all 8 targets with an average success rate of 74% ± 7% (sustained peak rate of 85%), with rapid convergence in the decoder parameters. Moreover, the subject is able to maintain high performance across 4 days with minimal adaptations to the decoder. Our results indicate that CLDA can be used to facilitate LFP-based BMI systems, allowing for both rapid improvement and maintenance of performance.

Volitional phase control of neural oscillations using a brain-machine interface

Strong oscillations present in local field potential (LFP) recordings at specific frequencies are thought to be correlated with behavior and involved in communication and coordination across brain areas. LFP-based brain-machine-interfaces (BMIs) lay the groundwork for using closed-loop BMIs to study oscillations across brain regions and to test hypotheses about the rigidity of relationships between oscillations. In this study, we demonstrate the first reported use of inter-regional phase differences to control a BMI. In a one-dimensional two target task where phase difference is mapped to cursor velocity, we show that a macaque monkey, implanted with microwire arrays bilaterally is able to control phase extracted from distal electrodes in left motor cortex (M1) and right dorsal pre-motor cortex (PMd) at frequencies of 20 Hz, 31 Hz, and 40 Hz well above chance, but not at 10 Hz and 50 Hz. Our results imply that within the beta band (15-40Hz) in motor cortex and pre-motor cortex, phase relationships can be volitionally controlled.

Assessing directed information as a method for inferring functional connectivity in neural ensembles

Neurons in the brain form complicated networks through synaptic connections. Traditionally, functional connectivity between neurons has been analyzed using simple metrics such as correlation, which do not provide direction of influence. Recently, an information theoretic measure known as directed information has been proposed as a way to capture directionality in the relationship, thereby moving towards a model of effective connectivity. This measure is grounded upon the concept of Granger causality and can be estimated by modeling neural spike trains as point process generalized linear models. However, the added benefit of using directed information to infer connectivity over conventional methods such as correlation is still unclear. Here, we propose a novel estimation procedure for the directed information. Using physiologically realistic simulations, we demonstrate that directed information can outperform correlation in determining connections between neural spike trains while also providing directionality of the relationship, which cannot be assessed using correlation.

Comparison of neural activity during closed-loop control of spike- or LFP-based brain-machine interfaces

Brain-machine interfaces (BMIs) have been developed using a variety of neural signals, including neuron action-potentials and local field-potentials (LFPs). However, little is known about the neural dynamics underlying closed-loop BMI control in these systems, and whether they might be shaped by the signal used for control. Better understanding the relationship between neural signals in closed-loop BMI could inform the design of future systems. We analyzed spiking and LFP activity in pre- and primary-motor cortices as a nonhuman primate performed closed-loop BMI driven by either spiking or LFP signals. Spike- and LFP-based BMI were done on different days. and all comparisons of activity are indirect. Both LFP and spiking activity showed significant task-related modulations in both types of BMI control. However, the neural dynamics varied with the control signal type. LFP signals, in particular, showed more directional modulation when BMI was controlled with LFPs.

Function Identification in Neuron Populations via Information Bottleneck

It is plausible to hypothesize that the spiking responses of certain neurons represent functions of the spiking signals of other neurons. A natural ensuing question concerns how to use experimental data to infer what kind of a function is being computed. Model-based approaches typically require assumptions on how information is represented. By contrast, information measures are sensitive only to relative behavior: information is unchanged by applying arbitrary invertible transformations to the involved random variables. This paper develops an approach based on the information bottleneck method that attempts to find such functional relationships in a neuron population. Specifically, the information bottleneck method is used to provide appropriate compact representations which can then be parsed to infer functional relationships. In the present paper, the parsing step is specialized to the case of remapped-linear functions. The approach is validated on artificial data and then applied to recordings from the motor cortex of a macaque monkey performing an arm-reaching task. Functional relationships are identified and shown to exhibit some degree of persistence across multiple trials of the same experiment.

Redundant information encoding in primary motor cortex during motor tasks

Information encoded in neuron ensembles has previously been hypothesized to be highly redundant, despite the apparent inefficiency of a redundant encoding system. The recent availability of intracortical, multi-electrode recordings has enabled the possibility of exploring how neuronal ensembles encode information as a whole. Applying concepts from information theory, we examined the redundancy of the target information encoded in both contralateral and ipsilateral hemispheres of the primary motor cortex (M1) in macaque monkeys performing a center-out reaching task. During movement, we reliably found neurons in contralateral M1 to contain higher target information and to be more redundant than ipsilateral M1 neurons, which is consistent with the conventional understanding that motor control is mainly governed by contralateral M1. Secondly, neuron ensembles in both hemispheres showed largely redundant information encoding. These results suggest that redundancy in information encoding is highly prevalent in the motor cortex and may contribute to proficient motor control.