Feedback error learning neural network for trans-femoral prosthesis.

Feedback-error learning (FEL) neural network was developed for control of a powered trans-femoral prosthesis. Nonlinearities and time-variations of the dynamics of the plant, in addition to redundancy and dynamic uncertainty during the double support phase of walking, makes conventional control methods very difficult to use. Rule-based control, which uses a knowledge base determined by machine learning and finite automata method is limited since it does not respond well to perturbations and environmental changes. FEL can be regarded as a hybrid control, because it combines nonparametric identification with parametric modeling and control. This paper presents simulation of a powered trans-femoral prosthesis controlled by a FEL neural network. Results suggest that FEL can be used to identify inverse dynamics of an arbitrary trans-femoral prosthesis during simple single joint movements (e.g., sinusoidal oscillations). The identified inverse dynamics then allows the tracking of an arbitrary trajectory such as a desired walking pattern within a multijoint structure. Simulation shows that the identified controller responds correctly when the leg motion is exposed to a perturbation such as a frequent change of the ground reaction force or the hip joint torque generated by the user. FEL eliminates the need for precise, tedious, and complex identification of model parameters.     

The Neurochip BCI: towards a neural prosthesis for upper limb function.

The Neurochip BCI is an autonomously operating interface between an implanted computer chip and recording and stimulating electrodes in the nervous system. By converting neural activity recorded in one brain area into electrical stimuli delivered to another site, the Neurochip BCI could form the basis for a simple, direct neural prosthetic. In tests with normal, unrestrained monkeys, the Neurochip continuously recorded activity of single neurons in primary motor cortex for several weeks at a time. Cortical activity was correlated with simultaneously-recorded electromyogram (EMG) activity from arm muscles during free behavior. In separate experiments with anesthetized monkeys, we found that microstimulation of the cervical spinal cord evoked movements of the arm and hand, often involving multiple muscles synergies. These observations suggest that spinal microstimulation controlled by cortical neurons could help compensate for damaged corticospinal projections.     

Automatic determination of synergies by radial basis function artificial neural networks for the control of a neural prosthesis.

This paper describes an automatic method for synthesizing the control for a neural prosthesis (NP) that could augment elbow flexion/extension and forearm pronation/supination in persons with hemiplegia. The basis for the control was a synergistic model of reaching and grasping that uses temporal and spatial synergies between the arm and body segments. The synergies were determined from the movement data measured in nondisabled persons during the performance of functional tasks. The work space was divided into six zones: distance (two attributes) and laterality (three attributes). Radial basis function artificial neural networks (RBF ANN) were used to determine synergies. Sets of RBF ANN characterized with good generalization were selected as control laws for elbow flexion/extension and forearm pronation/supination. The validation was performed for three categories: inter-subject, distance, and laterality generalization. For all of the defined spatial synergies, the correlation was high for inter-subject and distance, yet low for the laterality scenario. This suggests the necessity for implementing different maps for different directions, but the same maps for different distances. The natural movements of the upper arm then drive the lower arm (elbow flexion/extension and forearm pronation/supination) in a way that is very well suited for the administration of functional electrical therapy (FET) in persons with hemiplegia soon after the onset of impairment.     

Sensorimotor rhythm-based brain-computer interface (BCI): feature selection by regression improves performance.

People can learn to control electroencephalogram (EEG) features consisting of sensorimotor rhythm amplitudes and can use this control to move a cursor in one or two dimensions to a target on a screen. In the standard one-dimensional application, the cursor moves horizontally from left to right at a fixed rate while vertical cursor movement is continuously controlled by sensorimotor rhythm amplitude. The right edge of the screen is divided among 2-6 targets, and the user's goal is to control vertical cursor movement so that the cursor hits the correct target when it reaches the right edge. Up to the present, vertical cursor movement has been a linear function of amplitude in a specific frequency band [i.e., 8-12 Hz (mu) or 18-26 Hz (beta)] over left and/or right sensorimotor cortex. The present study evaluated the effect of controlling cursor movement with a weighted combination of these amplitudes in which the weights were determined by an regression algorithm on the basis of the user's past performance. Analyses of data obtained from a representative set of trained users indicated that weighted combinations of sensorimotor rhythm amplitudes could support cursor control significantly superior to that provided by a single feature. Inclusion of an interaction term further improved performance. Subsequent online testing of the regression algorithm confirmed the improved performance predicted by the offline analyses. The results demonstrate the substantial value for brain-computer interface applications of simple multivariate linear algorithms. In contrast to many classification algorithms, such linear algorithms can easily incorporate multiple signal features, can readily adapt to changes in the user's control of these features, and can accommodate additional targets without major modifications.     

Determining hearing threshold from brain stem evoked potentials.Optimizing a neural network to improve classification performance

Feed-forward neural networks in conjunction with back-propagation are an effective tool to automate the classification of biomedical signals. Most of the neural network research to date has been done with a view to accelerate learning speed. In the medical context, however, generalisation may be more important than learning speed. With the brain stem auditory evoked potential classification task described in this study, the authors found that parameter values that gave fastest learning could result in poor generalisation. In order to achieve maximum generalisation, it was necessary to fine tune the neural net for gain, momentum, batch size, and hidden layer size. Although this maximization could be time consuming, especially with larger training sets, the authors' results suggest that fine tuning parameters can have important clinical consequences, which justifies the time involved. In the authors' case, fine tuning parameters for high generalisation had the additional effect of reducing false negative classifications, with only a small sacrifice in learning speed     

Deep brain stimulation of the subthalamic nucleus improves set shifting but does not affect dual task performance in Parkinson's disease.

Problems with shifting attentional set and concurrent performance of tasks are key cognitive deficits in Parkinson's disease (PD). Our aim was to examine the effects of deep brain stimulation of the subthalamic nucleus on tests of set shifting and dual task performance in patients with PD. Twelve patients with PD were assessed on tests of set shifting and on dual task performance with subthalamic nucleus (STN) stimulation switched on and off in a counterbalanced order. All patients obtained a clinical benefit from deep brain stimulation (DBS) of the STN. STN stimulation significantly improved set shifting. The effect of DBS on dual task performance was not significant. Change in measures of set shifting was significantly associated with the change in the motor symptoms of PD with DBS. The improved set shifting with DBS of the STN in PD supports the critical role of the striato-frontal circuits in this cognitive function.