NEUROTECHNIX 2017 Abstracts


Area 1 - Neural Rehabilitation and Neuroprosthetics

Short Papers
Paper Nr: 3
Title:

Towards Developing a Brain-computer Interface for Automatic Hearing Aid Fitting based on the Speech-evoked Frequency Following Response

Authors:

Brian Heffernan, Hilmi R. Dajani and Christian Giguère

Abstract: One of the problems related to the use of hearing aids (HAs) is the difficulty in obtaining a best fit by adjusting different settings, such as those related to the gain and compression in different frequency bands. In this work, we propose using the information extracted from the brain’s frequency following response (FFR) to speech sounds to automatically adjust the settings of HAs via a brain-computer interface. This approach will be used to tune HAs to maximize the separation between the responses to different vowels and levels, with the aim of improving perceptual discrimination and loudness control in HA users.

Paper Nr: 13
Title:

Chronic Pain: Restoring the Central Nervous System “Body Image” with Virtual Reality

Authors:

Federica Alemanno, Elise Houdayer, Daniele Emedoli, Matteo Locatelli and Sandro Iannaccone

Abstract: n/a

Paper Nr: 17
Title:

A Biofeedback Control System of the Exoskeleton Trainer for Lower Limbs Motor Function Recovery

Authors:

Vasily Mironov, Innokentiy Kastalskiy, Sergey Lobov and Victor Kazantsev

Abstract: The development of robotic systems led to their spread to various spheres of human life. Medical rehabilitation is not an exception. It actively introduces robotic devices into medical practice, replacing traditional manual therapy with procedures using high-tech robotic devices. Robotic exoskeleton for rehabilitation, despite its long history of development, has become a technical innovation that has gained wide popularity in the medical environment in the last decade. Such devices are designed to compensate lower limb disability resulting from the brain or spinal cord injuries and diseases. In this paper, a new approach that uses neuromuscular signals from lower limbs to control the exoskeleton device is proposed. A method for detecting gait phases by the signal of EMG activity has been developed. We believe that control system presented will provide a more natural and effective method to mobilize patient resources and restore motor skills after stroke.

Area 2 - Neuroimaging and Neurosensing

Short Papers
Paper Nr: 6
Title:

An Attempt to Assess Alertness based on Emotions (From EEG Measures)

Authors:

Agnieszka Wolska, Dariusz Sawicki, Mariusz Wisełka and Szymon Ordysiński

Abstract: Alertness is conjugated to high level of awareness, low level of fatigue, and better mental performance. The level of alertness is related to melatonin level in the blood, which depends on light exposure. The analysis of EEG signal is one of the objective methods of alertness evaluation. Many publications have confirmed the possibility of effective and correct usage of Emotiv EPOC device in EEG signal recording. Emotiv Performance Metrics Detection Suite software registers real time changes in subjective emotions calculated from EEG signal. Two of them – engagement and excitement are associated with alertness. The article presents an attempt to assess the alertness level after exposure to different color of light based on EEG emotion recognition. 50 healthy subjects were recruited. Each subject took part in 3 experiments – each under different light scene (red, blue, white). Altogether 150 experiments with emotions registration (based on EEG) were carried out. The EEG registration was performed before the exposure and just after 40 minutes exposure to the particular light scene. Selected emotions have been used to assess the alertness state. The results of the experiments have been verified statistically. The results showed that the analysis of engagement and excitement changes gives the opportunity to assess alertness.

Area 3 - Neuroinformatics and Neurocomputing

Full Papers
Paper Nr: 7
Title:

Competition of Spike-Conducting Pathways in STDP Driven Neural Networks

Authors:

Sergey Lobov, Ksenia Balashova, Valeri A. Makarov and Victor Kazantsev

Abstract: Population spike or burst signaling is widely observed both in intact brain and neuronal cultures. Experimental evidence suggests that locally applied electrical stimuli can shape the network architecture and thus the neuronal response. However, there is no clue on how this process can be controlled. In this work we study a realistic model of a culture of cortical-like neurons with spike timing dependent plasticity (STDP). We show that the network dynamics is driven by a competition of spike-conducting pathways, which influences the learning ability of the network. Even in the case of single-electrode stimulation the network dynamics can be complex. Self-establishing spike-conducting pathways, different from those we expect to strengthen, can interfere the process of the network structuring. It leads to an intermittent regime: the time intervals of well-pronounced population spikes synchronized with the stimulus are alternated by intervals of asynchronous dynamics. Under multi-electrode stimulation of an unstructured network the competition of spike-conducting pathways destroys the unconditional learning. The STDP stimulation protocol fails to work at the network scale. To overcome this restriction we propose to use structured neural network and show that one can train such a network and achieve spiking activity circulating in the network after the stimulus has been switched off.

Paper Nr: 10
Title:

Probabilistic Symbol Encoding for Convolutional Associative Memories

Authors:

Igor Peric, Alexandru Lesi, Daniel Spies, Stefan Ulbrich, Arne Roennau, Marius Zoellner and Ruediger Dillman

Abstract: Vector Symbolic Architectures (VSAs) define a set of operations for association, storage, manipulation and retrieval of symbolic concepts, represented as fixed-length vectors in IRn. A specific instance of VSAs, Holo- graphic Reduced Representations (HRRs), have proven to exhibit properties similar to human short-term mem- ory and as such are interesting for computational modelling. In this paper we extend the HRR approach by introducing implicit, topology-preserving encoding and decoding procedures. We propose to replace unique symbolic representations with symbols based on probability density functions. These symbols must be ran- domly permuted to ensure the uniform distribution of signals across Fourier space where embedding takes place. These novel encoding schemes eliminate the need for so-called clean-up modules after memory re- trieval (e.g., self-organizing maps). Effectively each encoding implicitly represents its scalar symbol, so no further lookup is needed. We further show that our encoding scheme has a positive impact on memory capacity in comparison to the original capacity benchmark for HRRs (Plate, 1995). We also evaluate our memories in two different robotics tasks: visual scene memory and state machine scripting (holographic controllers).

Short Papers
Paper Nr: 8
Title:

Application of Artificial Neural Models for Planning Sport Training in 110m Hurdles

Authors:

Krzysztof Przednowek, Janusz Iskra, Tomasz Krzeszowski and Karolina H. Przednowek

Abstract: This paper presents the use of artificial neural networks for planning sport training in 110 meters hurdles. The model was calculated based on the training data of the Polish National Team hurdlers. The analysis was based on 120 training plans that represent a different period in the annual training cycle. The MLP and RBF networks were used in this study as a predictive model. The neural network developed has 6 inputs representing the parameters of the athlete and 15 outputs representing the training loads. To evaluate the model, a set of 20 records were used. The smallest prediction error was obtained for the multilayer perceptron with 9 neurons in the hidden layer and a hyperbolic tangent as the activation functions. The resulting model may be used as a tool to assist coaches in planning training loads during the selected training period.

Paper Nr: 11
Title:

Semi-Supervised Spiking Neural Network for One-Shot Object Appearance Learning

Authors:

Igor Peric, Robert Hangu, Jacques Kaiser, Stefan Ulbrich, Arne Roennau, Johann Marius Zoellner and Ruediger Dillman

Abstract: We present a network of spiking neurons which extracts intermediate-level features from a set of classes in an unsupervised manner and uses them to later learn new, unrelated classes with just a few training examples, also called one-shot learning. The framework is built on the biologically plausible neurosimulator NEST developed and used by neuroscientists, giving the work an unprecedented biological plausibility over previous similar approaches which use custom-built systems tuned to their needs. Furthermore, the learning of the classes happens in a continuous manner, without scripted interruptions and external interventions to the neuron states during simulation, which draws this work even closer to biological realism. The high quality of the learned features is confirmed by achieving a close to state-of-the-art F1 score of 97% during the recognition of the same classes, while obtaining a score as high as 72% for one-shot learning. This paper focuses more on the biological plausibility of the presented ideas and less on the concrete object classification mechanisms.

Paper Nr: 19
Title:

Holistic Model of Cognitive Limbs for Dynamic Situations

Authors:

C. Calvo, I. Kastalskiy, J. A. Villacorta-Atienza, M. Khoruzhko and V. A. Makarov

Abstract: Cognitive object handling and manipulation are vital skills for humans and future humanoid robots. However, the fundamental bases of how our brain solves such tasks remain largely unknown. Here we provide a novel approach that describes the problem of limb movements in dynamic situations on an abstract cognitive level. The approach involves two main steps: i) Transformation of the problem from the limb workspace to the so-called handspace, which represents the limb as a point and obstacles as objects of complex shapes, and ii) Construction of a generalized cognitive map (GCM) in the handspace by a neural network simulating activation wave. The GCM enables tracing a trajectory to a target that can be followed by the limb, which ensures collision-free movement and target catching in the workspace. We validate our approach by numerical simulations on an avatar developed for a humanoid robot Poppy.

Area 4 - Neuromodulation and Neural Engineering

Short Papers
Paper Nr: 9
Title:

Multi-Modal Integrated Mini-QEEG Solution with Results, Training Protocols and Neurofeedback in Real-time

Authors:

Francisco Marques-Teixeira, Horácio Tomé-Marques, João Andrade and João Marques-Teixeira

Abstract: Many Neurofeedback softwares allow clinicians to develop modular protocols. Some even allow to be performed quantitative Electroencephalography (QEEG) analysis. However, an all-in-one solution does not exist, with built-in decision tree, that permits to perform a QEEG analysis, apply a protocol decision, and perform subsequent iterated mini-QEEG analysis, and subsequente protocol decisions and training, in the same system. Our application, besides being able to make the identification of the brain dysfunctional patterns concerning its electrophysiology, and accurately choose the Neurofeedback training protocols to apply, performs real-time Neurofeedback. We compared our system with a conventional EEG and QEEG system in a proof of concept rational, obtaining an average Pearson Correlation Coefficient of 0.89 regarding the dysfunctional patterns and protocols, as well as remitted dysfunctional patterns. In conclusion, our application is pervasive, scalable and potentially ubiquitous. And it can be extended into multiple and different consumer fields.

Paper Nr: 15
Title:

In-vitro Modeling of Electrode-tissue Parameters

Authors:

Maran Ma and Timothy E. Kennedy

Abstract: The long-term causes underlying the failure of neural recording electrodes is an active question in the community of neural implant users and developers. It is known that a variety of phenomena contribute to signal degradation, but to engineer better devices, the impact of each factor must be quantified and prioritized. The causes of gradual loss of signal fidelity on individual electrodes include: 1) insulation breakdown, 2) biofouling, 3) glial ensheathment, and 4) neuronal death. One challenge to teasing these factors apart is that they occur simultaneously in-vivo. Another is that impedance, the primary method of monitoring electrode status, is affected by multiple factors. The purpose of this project is to build a simple and cost effective in-vitro setup to model each phenomenon individually, and quantify its impact on both impedance spectra and electrophysiological recording quality.