D.J. Weber

Work toward real-time control of a cortical neural prothesis

Implantable devices that interact directly with the human nervous system have been gaining acceptance in the field of medicine since the 1960s. More recently, as is noted by the FDA approval of a deep brain stimulator for movement disorders, interest has shifted toward direct communication with the central nervous system (CNS). Deep brain stimulation (DBS) can have a remarkable effect on the lives of those with certain types of disabilities such as Parkinsons disease, Essential Tremor, and dystonia. To correct for many of the motor impairments not treatable by DBS (e.g. quadriplegia), it would be desirable to extract from the CNS a control signal for movement. A direct interface with motor cortical neurons could provide an optimal signal for restoring movement. In order to accomplish this, a real-time conversion of simultaneously recorded neural activity to an online command for movement is required. A system has been established to isolate the cellular activity of a group of motor neurons and interpret their movement-related information with a minimal delay. The real-time interpretation of cortical activity on a millisecond time scale provides an integral first step in the development of a direct brain-computer interface (BCI).

The role of vision on hand preshaping during reach to grasp

During reaching to grasp objects with different shapes hand posture is molded gradually to the objects contours. The present study examined the extent to which the temporal evolution of hand posture depends on continuous visual feedback. We asked subjects to reach and grasp objects with different shapes under five vision conditions (VCs). Subjects wore liquid crystal spectacles that occluded vision at four different latencies from onset of the reach. As a control, full-vision trials (VC5) were interspersed among the blocked vision trials. Object shapes and all VCs were presented to the subjects in random order. Hand posture was measured by 15 sensors embedded in a glove. Linear regression analysis, discriminant analysis, and information theory were used to assess the effect of removing vision on the temporal evolution of hand shape. We found that reach duration increased when vision was occluded early in the reach. This was caused primarily by a slower approach of the hand toward the object near the end of the reach. However, vision condition did not have a significant effect on the covariation patterns of joint rotations, indicating that the gradual evolution of hand posture occurs in a similar fashion regardless of vision. Discriminant analysis further supported this interpretation, as the extent to which hand posture resembled object shape and the rate at which hand posture discrimination occurred throughout the movement were similar across vision conditions. These results extend previous observations on memory-guided reaches by showing that continuous visual feedback of the hand and/or object is not necessary to allow the hand to gradually conform to object contours.

Adaptive behavior of cortical neurons during a perturbed arm-reaching movement in a nonhuman primate

This chapter provides evidence of spatial and temporal changes in the behavior of neurons within Areas 5 and 4 of the sensorimotor cortex of a nonhuman primate while it was executing a perturbed arm-reaching task. Chronically implanted electrode arrays were used to record simultaneously from 37 to 58 neurons. Also measured were the trajectory of arm movement, EMG activity in selected arm muscles and the perturbation force applied to the arm. The adaptation in Area 4 neurons behavior usually involved a reduction in the latency from the onset of the perturbation to the peak-firing rate of the cell. In contrast, Area 5 neurons exhibited no such adaptive change in this latency. In each cortical area, the adaptation was not uniform across all neurons, and the spatial pattern of neuronal population behavior changed over the period of behavioral adaptation. We also found that the direction of arm movement and its configuration were important in determining which control strategy (predictive trajectory compensation or stiffness control) the animal used to overcome the externally applied perturbation for an improved performance of the reaching task.

Disassociation between primary motor cortical activity and movement kinematics during adaptation to reach perturbations

The relationship between movement kinematics and motor cortical activity was studied in monkeys performing a center-out reaching task during their adaptation to force perturbations applied to the wrist. The main feature of adaptive changes in movement kinematics was anticipatory deviation of hand paths in the direction opposite to that of the upcoming perturbation. We identified a group of neurons in the dorsal lateral portion of the primary motor cortex where a gradual buildup of spike activity immediately preceding the actual (in perturbation trials) or the would-be (in unperturbed/catch trials) perturbation onset was observed. These neurons were actively involved in the adaptation process, which was evident from the gradual increase in the amplitude of their movement-related modulation of spike activity from virtual zero and development of certain directional tuning pattern (DTP). However, the day-to-day dynamics of the kinematics adaptation was dramatically different from that of the neuronal activity. Hence, the adaptive modification of the motor cortical activity is more likely to reflect the development of the internal model of the perturbation dynamics, rather than motor instructions determining the adaptive behavior.

Adaptation of arm movement control under repeated perturbations

To develop control strategies for a prosthetic arm or electrical stimulation of paralyzed muscles, we have designed a novel approach to investigate how the neural system utilizes available sensory information to learn and adapt to a perturbation for the intended arm reaching movement. A parallel approach of both human and animal experiments allows us to collect data from cortical neurons, muscles and arm movement kinematics to analyze coordinated changes in control strategy of various levels in the sensorimotor system. We simultaneously record neural activities in motor cortex, EMG responses in 7 arm muscles, and arm movement trajectories during a visually guided reaching task. Force perturbations (an impulse of 75-100 ms) are delivered through a string attached to the wrist of the moving arm. Preliminary results and data analysis demonstrate that human subjects develop an anticipatory strategy to improve performance of reaching tasks under repeated force perturbations applied to the wrist. The same experimental protocol has been modified to rhesus monkeys with implanted multichannel cortical electrodes. Data collected from this experiment show that correlation of adaptation among individual and populations of cortical neurons, muscle coordination patterns, and kinematics can be established.

Adaptation in cortical control of arm movement

Abstract By simultaneously recording multi-neuronal activities through chronically implanted electrode arrays in motor and sensory cortices, arm movement trajectory, and muscle activities, we observed both spatial and temporal differences in neuronal activities during different phases of a reaching task. We also discovered that when developing an effective control strategy to overcome an externally applied perturbation, the direction of the movement and the arm configuration played an important role in determining which control strategy to apply. The results indicated that predictive trajectory compensation was often adopted while stiffness control was also utilized.

Response of simultaneously recorded cortical neurons to perturbations in arm trajectory

Cortical neural prostheses may someday be used to augment or restore lost motor functions by deriving movement commands from neural activity recorded in the brain. Many studies have identified relationships between the discharge patterns of cortical neurons and arm movement parameters (e.g. direction and force). We are currently investigating these relationships in simultaneously recorded cortical neurons to generate a movement command in real-time. A perturbation paradigm has been implemented to study changes in cortical activity as the monkey responds and adapts to a perturbation in trajectory. Over 40 neurons were recorded simultaneously over a 2-week period. Distinct and stable cortical responses (to movement and perturbation) were observed across the two weeks of trials.

Modifications of motor cortical activity induced by adaptation to movement perturbations as revealed by chronic multielectrode recordings in monkeys

Knowledge of the properties of motor cortical neurons is crucial for solving the problem of developing a flexible and robust brain-computer interface (BCI). The results of chronic multielectrode recording from the primary motor cortical area of monkey brain during the animals performance of a center-out 3D reaching task and adaptation to external force perturbation of the movement are described. Three different types of learning-related modifications of cell spike activity have been identified and described. The most important feature of the first one is the development of anticipatory spike activity preceding the onset of the perturbation. The second and third types are related to the gradual increase of spike activity modulation with respect to the reaching movement and the directional differentiation of this activity reflected in its progressive dependence on the coordinates of the reach target. Adaptive changes of the last type were observed in the rostral-medial part of the primary motor cortex. The capacity of those neurons for this type of adaptation is crucially important for the possibility of their utilization for decoding the direction of subjects intended movement. It also indicates that the subjects capability to use a BCI can be improved with training.

Human Arm Trajectory: Planning, Execution, and Control

Abstract A three dimensional, five degrees of freedom human arm model has been developed to study the planning, execution and control of hand movement. The hand trajectory is assumed to derive from the population vector of neural cells of motor cortex. The real time calculation of joint trajectories from the hand trajectory is based on a biologically inspired algorithm. The instantaneous angular velocity for each joint is determined by the cross product of two vectors: the joint vector and the error vector. A nonlinear gain modulation adjusts the gain for each degree of freedom according to the instantaneous joint position relative to its physiological limit. The algorithm does not use inverse kinematics to determine joint trajectories from the hand position, and can be adapted to situations where the final target position is moving, for example, to track or to intercept a moving object. Several simulation results are presented to illustrate the merit of the algorithm.