In this work, we present a set of methods aimed at improving the discriminative power of time-varying features of signals that contain noise. These methods use properties of noise signals as well as information theoretic techniques to factor types of noise and support signal inference for electroencephalographic (EEG) based brain-machine interfaces (BMI). EEG data were collected over two studies aimed at addressing Psychophysiological issues involving symmetry and mental rotation processing. The Psychophysiological data gathered in the mental rotation study also tested the feasibility of using dissociations of mental rotation tasks correlated with rotation angle in a BMI. We show the feasibility of mental rotation for BMI by showing comparable bitrates and recognition accuracy to state-of-the-art BMIs. The conclusion is that by using the feature selection methods introduced in this work to dissociate mental rotation tasks, we produce bitrates and recognition rates comparable to current BMIs.

Thesis (M.S.)--Mississippi State University. Department of Computer Science. / Title from title screen. Includes bibliographical references.

Thesis (M.S.)--Mississippi State University. Department of Computer Science and Engineering. / Title from title screen. Includes bibliographical references.

The document describes the development of a micro-processor based terminal multiplexer to connect four terminals to a standard Hewlett Packard series 1000 mini-computer. The project was required to fulfill the dual roll of both increasing the number of terminals that the HPI000 could support and of reducing the peripheral load on the host CPU. The final product occupied a standard 200mm square HP size interface card and used an 8085 micro-processor and several 8085 family peripheral chips to provide four full duplex serial channels and a high speed data link with the host. A multi-tasking executive was written to control the multiplexer software which was finally implemented as 15 independent tasks occupying 8 kilo-bytes of eprom. The software was written to perform all terminal interaction and editing in order to reduce the host CPU involvement to a single interrupt per record. The resultant interface proved capable of handling an aggregate throughput in excess of 4000 characters per second which was sufficient to cope with all four terminals running at 9600 bits per second, even when all four were transferring in burst mode. The interface also proved to be between five and eighteen times less demanding on the host than the two standard Hewlett Packard interfaces then available. When compared to the low cost HP12531 interface, the multiplexer increased the 9600b/s terminal handling capability of the host from 3 terminals to 52. / Thesis (M.Sc.-Electronic Engineering)-University of Natal, 1987.

Thesis (Ph. D.)--State University of New York at Binghamton, Computer Science Department, 2006. / Includes bibliographical references.

Thesis (Ph.D.)--University of Central Florida, 2008. / Advisers: Ratan K. Guha, Joohan Lee. Includes bibliographical references (p. 132-143).

The development of effective multidimensional map-based interfaces is an important area of research in need of design techniques and guidelines. To date, guidelines for multidimensional interfaces have been generalized from text-based interfaces and few experimental studies have been conducted to asses their effectiveness. Guidelines for design were studied with the goal of extending the current body of knowledge about the usability of these interfaces. Based on design guidelines, multidimensional map-based interfaces with various levels of depth and breath, with and without scent-based components were used to perform simple and compound tasks. The goal of this study was to investigate the effectiveness of design guidelines on response time, preferences, and navigation and task accuracy. Results showed relationships exist among navigation and task accuracy, response time, and preferences within simple or compound tasks. However, few relationships exist between simple and compound tasks. Contrary to results from previous research, interface depth and breadth was found to have no significant effect on navigation and task accuracy or response time. For compound tasks, interfaces with scent-based components were found to be more effective regarding task accuracy at greater depth levels. The absence of scent in the interface was shown to be more efficient regarding response time and navigation accuracy during compound tasks.

Leading human-computer interaction (HCI) researchers recognize a fundamental difference exists between men and women. Some HCI research has been done regarding gender differences in hand-eye coordination for interactions with computer touch display interfaces, navigation through virtual environments (VE) and language in computer-mediated communication. In these previous studies, gender differences were found in the use of words and language in computer-mediated communication and in navigation strategies for VE but no gender-related differences were found for the hand-eye coordination needed to effectively use a touch display. The current study used a cross-modal (auditory-visual), dual-task, computer interface to examine gender differences in crisis control simulations. For the primary task of alarm monitoring, no gender differences were found for average or maximum response and completion times. Likewise, no gender differences were found in terms of error rates for the primary task or the number correct on the secondary task. However, in terms of minimum response and completion times for alarm monitoring, gender differences were found.

As Interfaces Cérebro-Computador (do inglês Brain-Computer Interfaces BCI) são sistemas que visam permitir a interação entre usuários e máquinas por meio do monitoramento das atividades cerebrais. Sistemas de BCI são considerados como uma alternativa para que pessoas com perda severa ou total do controle motor, tais como as que sofrem de Esclerose Lateral Amiotrófica, possam contar com algum controle sobre o ambiente externo. Para mapear intenções individuais em operações de máquina, os sistemas de BCI empregam um conjunto de etapas que envolvem a captura e pré-processamento dos sinais cerebrais, a extração e seleção de suas características mais relevantes e a classificação das intenções. O projeto e a implementação de sistemas de BCI viáveis ainda são questões em aberto devido aos grandes desafios encontrados em cada uma de suas etapas. Esta lacuna motivou este trabalho de mestrado o qual apresenta uma avaliação dos principais extratores de características utilizados para classificar ensaios de imagética motora, cujos dados foram obtidos por meio de eletroencefalografia (EEG) e apresentam influências de artefatos, mais precisamente daqueles produzidos por interferências provenientes de atividades oculares (monitoradas por eletrooculografia EOG). Foram considerados sinais coletados pela BCI Competition IV-2b, os quais contêm informações sobre três canais de EEG e três outros de EOG. Como primeira etapa, foi realizado o pré-processamento desses canais utilizando a técnica de Análise de Componentes Independentes (ICA) em conjunto com um limiar de correlação para a remoção de componentes associados a artefatos oculares. Posteriormente, foram avaliadas diferentes abordagens para a extração de características, a mencionar: i) Árvore Diádica de Bandas de Frequências (ADBF); ii) Padrões Espaciais Comuns (CSP); iii) Padrões Espectro-Espaciais Comuns (CSSP); iv) Padrões Esparsos Espectro-Espaciais Comuns (CSSSP); v) CSP com banco de filtros (FBCSP); vi) CSSP com banco de filtros (FBCSSP); e, finalmente, vii) CSSSP com banco de filtros (FBCSSSP). Contudo, como essas técnicas podem produzir espaços de exemplos com alta dimensionalidade, considerou-se, também, a técnica de Seleção de Características baseada em Informação Mútua (MIFS) para escolher os atributos mais relevantes para o conjunto de dados adotado na etapa de classificação. Finalmente, as Máquinas de Vetores de Suporte (SVM) foram utilizadas para a classificação das intenções de usuários. Experimentos permitem concluir que os resultados do CSSSP e FBCSSSP são equiparáveis àqueles produzidos pelo estado da arte, considerando o teste de significância estatística de Wilcoxon bilateral com confiança de 0, 95. Apesar disso o CSSSP tem sido negligenciado pela área devido ao fato de sua parametrização ser considerada complexa, algo que foi automatizado neste trabalho. Essa automatização reduziu custos computacionais envolvidos na adaptação das abordagens para indivíduos específicos. Ademais, conclui-se que os extratores de características FBCSP, CSSP, CSSSP, FBCSSP e FBCSSSP não necessitam da etapa de remoção de artefatos oculares, pois efetuam filtragens por meio de modelos autoregressivos. / Brain-Computer Interfaces (BCI) employ brain imaging to enable human-machine interaction without physical control. BCIs are an alternative so that people suffering from severe or complete loss of motor control, like those with Amyotrophic Lateral Sclerosis (ALS), may have some interaction with the external environment. To transform individual intentions onto machine operations, BCIs rely on a series of steps that include brain signal acquisition and preprocessing, feature extraction, selection and classification. A viable BCI implementation is still an open question due to the great challenges involved in each one of these steps. This gap motivated this work, which presents an evaluation of themain feature extractors used to classify Motor Imagery trials, whose data were obtained through Electroencephalography (EEG) influenced by ocular activity, monitored by Electrooculography (EOG). In this sense, signals acquired by BCI Competition IV-2b, were considered. As first step the preprocessing was performed through Independent Component Analysis (ICA) together with a correlation threshold to identify components associated with ocular artifacts. Afterwards, different feature extraction approaches were evaluated: i) Frequency Subband Dyadic Three; ii) Common Spatial Patterns (CSP); iii) Common Spectral-Spatial Patterns (CSSP); iv) Common Sparse Spectral-Spatial Patterns (CSSSP); v) Filter Bank Common Spatial Patterns (FBCSP); vi) Filter Bank Common Sectral-Spatial Patterns (FBCSSP); and, finally, vii) Filter Bank Sparse Spectral- Spatial Patterns (FBCSSSP). These techniques tend to produce high-dimensional spaces, so a Mutual Information-based Feature Selection was considered to select signal attributes. Finally, Support Vector Machines were trained to tackle the Motor Imagery classification. Experimental results allow to conclude that CSSSP and FBCSSSP are statistically equivalent the state of the art, when two-sided Wilcoxon test with 0, 95 confidence is considered. Nevertheless, CSSSP has been neglected by this area due to its complex parametrization, which is addressed in this work using an automatic approach. This automation reduced computational costs involved in adapting the BCI system to specific individuals. In addition, the FBCSP, CSSP, CSSSP, FBCSSP and FBCSSSP confirm to be robust to artifacts as they implicitly filter the signals through autoregressive models.

Esta dissertação de mestrado tem por objetivo analisar os conjuntos-K, uma hierarquia de redes neurais biologicamente mais plausíveis, e aplicá-los ao problema de classificação de imagética motora através do eletroencefalograma (EEG). A imagética motora consiste no ato de processar um movimento motor da memória humana de longo tempo para a memória de curto prazo. A imagética motora deixa um rastro no sinal do EEG que torna possível a identificação e classificação dos diferentes movimentos motores. A tarefa de classificação de imagética motora através do EEG é reconhecida como complexa devido à não linearidade e quantidade de ruído da série temporal do EEG e da pequena quantidade de dados disponíveis para aprendizagem. Os conjuntos-K são um modelo conexionista que simula o comportamento dinâmico e caótico de populações de neurônios do cérebro e foram modelados com base em observações do sistema olfatório feitas por Walter Freeman. Os conjuntos-K já foram aplicados em diversos domínios de classificação diferentes, incluindo EEG, tendo demonstrado bons resultados. Devido às características da classificação de imagética motora, levantou-se a hipótese de que a aplicação dos conjuntos-K na tarefa pudesse prover bons resultados. Um simulador para os conjuntos-K foi construído para a realização dos experimentos. Não foi possível validar a hipótese levantada no trabalho, dado que os resultados dos experimentos realizados com conjuntos-K e imagética motora não apresentaram melhorias significativas para a tarefa nas comparações realizadas. / This dissertation aims to examine the K-sets, a hierarchy of biologically plausible neural networks, and apply them to the problem of motor imagery classification through electroencephalogram (EEG). Motor imagery is the act of processing a motor movement from long-term to short-term memory. Motor imagery leaves a trail in the EEG signal, which makes possible the identification and classification of different motor movements. Motor imagery classification is a complex problem due to non-linearity of the EEG time series, low signal-to-noise ratio, and the small amount of data typically available for learning. K-sets are a connectionist model that simulates the dynamic and chaotic behavior of populations of neurons in the brain, modeled based on observations of the olfactory system by Walter Freeman. K-sets have already been used in several different classification domains, including EEG, showing good results. Due to the characteristics of motor imagery classification, a hypothesis that the application of K-sets in the task could provide good results was raised. A simulator for K-sets was created for the experiments. Unfortunately, the hypothesis could not be validated, as the results of the conducted experiments with K-sets and motor imagery showed no significant improvements in comparison in the task performed.