International IEEE/EMBS Conference on Neural Engineering : [proceedings]. International IEEE EMBS Conference on Neural Engineering最新文献

Electrical stimulation using non-periodic biomimetic stimulation pattern has been shown to be effective in various critical biomedical applications. However, the existing programmable stimulators that support this protocol are non-portable and have architectures that are not translatable to wearable or implantable applications. In this work, we present a 32-channel neural stimulator system based on an implantable System-On-Chip (SoC) that addresses these technological challenges. The system is designed to be portable, powered by a single battery, wirelessly controlled, and versatile to perform concurrent multi-channel stimulation with independent arbitrary waveforms. The experimental results demonstrate multi-channel stimulation mimicking electromyography (EMG) waveforms and randomly-spaced stimulation pulses mimicking neuronal firing patterns. This compact and highly flexible prototype can support various neuromodulation researches and animal studies and serves as a precursor for the development of the next generation implantable biomimetic stimulator.

A major issue with upper limb prostheses is the disconnect between sensory information perceived by the user and the information perceived by the prosthesis. Advances in prosthetic technology introduced tactile information for monitoring grasping activity, but visual information, a vital component in the human sensory system, is still not fully utilized as a form of feedback to the prosthesis. For able-bodied individuals, many of the decisions for grasping or manipulating an object, such as hand orientation and aperture, are made based on visual information before contact with the object. We show that inclusion of neuromorphic visual information, combined with tactile feedback, improves the ability and efficiency of both able-bodied and amputee subjects to pick up and manipulate everyday objects. We discovered that combining both visual and tactile information in a real-time closed loop feedback strategy generally decreased the completion time of a task involving picking up and manipulating objects compared to using a single modality for feedback. While the full benefit of the combined feedback was partially obscured by experimental inaccuracies of the visual classification system, we demonstrate that this fusion of neuromorphic signals from visual and tactile sensors can provide valuable feedback to a prosthetic arm for enhancing real-time function and usability.

Understanding how different brain areas interact to generate complex behavior is a primary goal of neuroscience research. One approach, functional connectivity analysis, aims to characterize the connectivity patterns within brain networks. In this paper, we address the problem of discriminative connectivity, i.e. determining the differences in network structure under different experimental conditions. We introduce a novel model called Sparse Multi-task Inverse Covariance Estimation (SMICE) which is capable of estimating a common connectivity network as well as discriminative networks across different tasks. We apply the method to EEG signals after solving the inverse problem of source localization, yielding networks defined on the cortical surface. We propose an efficient algorithm based on the Alternating Direction Method of Multipliers (ADMM) to solve SMICE. We apply our newly developed framework to find common and discriminative connectivity patterns for α-oscillations during the Sleep Onset Process (SOP) and during Rapid Eye Movement (REM) sleep. Even though both stages exhibit a similar α-oscillations, we show that the underlying networks are distinct.

Task-induced variations in neural activity and their effects on the topological architecture of intrinsic connectivity networks (ICNs) of the brain are still a matter of ongoing research. In this exploratory study, we used spatial independent component analysis (ICA) as a data-driven technique to characterize ICNs related to two different tasks in healthy subjects who underwent 3T blood oxygen level-dependent (BOLD) functional magnetic resonance imaging (fMRI). The fMRI tasks consisted of (a) a viscerosensory stimulation of an internal organ (interoceptive task), and (b) passive viewing of emotionally expressive faces and pictures from the International Affective Picture System (exteroceptive emotion task). Comparison of the network volumes and peak activations during each task condition demonstrated that changes in ICN volume and corresponding peak activation differed between the interoceptive and exteroceptive emotion tasks when compared to the baseline rest. Further, salience network was the most task-activated ICN for both fMRI task conditions. However, different spatial characteristics were observed between the salience networks derived from the interoceptive task and the one derived from the exteroceptive emotion task. This study is a step in the direction of better understanding the influence of task condition on ICN topology. Future research with a larger sample size and task variations should delve deeper into what aspects of network topology really matter, with further investigations regarding the observed differences due to gender and age.

Current knowledge of coordinated motor control of multiple muscles is derived primarily from invasive stimulation-recording techniques in animal models. Similar studies are not generally feasible in humans, so a modeling framework is needed to facilitate knowledge transfer from animal studies. We describe such a framework that uses a deep neural network model to map finite element simulation of transcranial magnetic stimulation induced electric fields (E-fields) in motor cortex to recordings of multi-muscle activation. Critically, we show that model generalization is improved when we incorporate empirically derived physiological models for E-field to neuron firing rate and low-dimensional control via muscle synergies.

This study proposes a novel approach for evaluating the task invariance of muscle synergies, vital for potential implementation in improving prosthetic hand control. We do this by using a transfer learning paradigm to test for invariance across a relatively small set of hand/forearm muscle synergies, derived from electromyographic (EMG) activation patterns during voluntary behaviors such as finger spelling and grasp mimicking postures and unconstrained exploration. EMG for each task were decomposed using non-negative matrix factorization into synergy and weight matrices, and cross-task weights for each task were then reconstructed by employing the base matrices from different tasks. Support Vector Machine and Extreme Learning Machine classifiers were used to classify the resulting weights in order to compare their performance, as well as their behaviors as a function of synergy rank. Both algorithms showed robust and significantly higher performance, compared to two distinct randomized controls, with lower rank EMG representations, both within and between tasks/postures, supporting hypotheses of functional invariance of multi-muscle synergies. Our results suggest that this invariance could be leveraged to efficiently calibrate postures for prosthetic hand implementation by transferring learned EMG patterns from unconstrained movements to other tasks.

In this paper, we present the design and performance of a portable, arbitrary waveform, multichannel constant current electrotactile stimulator that costs less than $30 in components. The stimulator consists of a stimulation controller and power supply that are less than half the size of a credit card and can produce ±15 mA at ±150 V. The design is easily extensible to multiple independent channels that can receive an arbitrary waveform input from a digital-to-analog converter, drawing only 0.9 W/channel (lasting 4-5 hours upon continuous stimulation using a 9 V battery). Finally, we compare the performance of our stimulator to similar stimulators both commercially available and developed in research.

With the advent of genetically-encoded optical tools to trigger or report neuronal activity, new designs for multielectrode arrays (MEAs) used in neural interfacing incorporate both optical and electrical modes of stimulating or recording neural activity. Likewise, the need to improve upon the biocompatibility of implanted MEAs has moved the field towards the use of softer, more compliant materials in device fabrication. However, there is limited available information on the impact of the materials used in MEAs on the function of interfaced individual neurons and neuronal networks. We assessed the responses of rat cortical neurons on optically transparent materials commonly used in the construction of "next-generation" devices: indium tin oxide (ITO), parylene-C, and polydimethylsiloxane (PDMS). We found that neuronal network formation and spiking responses to electrical stimulation were enhanced in neurons cultured on ITO. We observed reduced excitability and synaptic connectivity between neurons cultured on PDMS. We hypothesize that the superior conductivity of ITO and suboptimal neuronal attachment to PDMS contributed to our results.

Current Brain Computer Interface (BCI) systems are limited by relying on neuronal spikes and decoding limited to kinematics only. For a BCI system to be practically useful, it should be able to decode brain information on a continuous basis with low latency. This study investigates if force can be decoded from local field potentials (LFP) recorded with deep brain electrodes located at the Subthalamic nucleus (STN) using data from 5 patients with Parkinson's disease, on a continuous basis with low latency. A Wiener-Cascade (WC) model based decoder was proposed using both time-domain and frequency-domain features. The results suggest that high gamma band (300-500Hz) activity, in addition to the beta (13-30Hz) and gamma band (55-90Hz) activity is the most informative for force prediction but combining all features led to better decoding performance. Furthermore, LFP signals preceding the force output by up to 1256 milliseconds were found to be predictive of the force output.