The last decade has seen a surge in the development of Brain-Machine Interfaces (BMI) as assistive neural devices for paralysis patients. Current BMI research typically involves a subject performing movements by controlling a robotic prosthesis. The neural signal that we consider for analysis is the Local Field Potential (LFP). The LFP is a low frequency neural signal recorded from intra-cortical electrodes, and has been recognized as one containing movement information. This thesis investigates hand-movement prediction using LFP data as input. In Chapter 1, we give an overview of Brain Machine Interfaces. In Chapter 2, we review the necessary concepts in time series analysis and pattern recognition. In the final chapter, we discuss classification accuracies when considering Summed power and Coherence as feature vectors.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19126号 / 情博第572号 / 新制||情||100(附属図書館) / 32077 / 京都大学大学院情報学研究科システム科学専攻 / (主査)教授 石井 信, 教授 田中 利幸, 教授 加納 学 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

A brain-machine interface (BMI) is a system that allows the communication between the central nervous system (CNS) and an external device (Wolpaw et al. 2002). Applications of BMIs include the control of external prostheses, cursors and spellers, to name a few. The BMIs developed by various research groups differ in their characteristics (e.g. continuous or discrete, synchronous or asynchronous, degrees of freedom, others) and, in spite of several initiatives towards standardization and guidelines, the cross comparison across studies remains a challenge (Brunner et al. 2015; Thompson et al. 2014). Here, we used a 64-channel EEG equipment to acquire data from 19 healthy participants during three different tasks (SSVEP, P300 and hybrid) that allowed four choices to the user and required no previous neurofeedback training. We systematically compared the offline performance of the three tasks on the following parameters: a) accuracy, b) information transfer rate, c) illiteracy/inefficiency, and d) individual preferences. Additionally, we selected the best performing channels per task and evaluated the accuracy as a function of the number of electrodes. Our results demonstrate that the SSVEP task outperforms the other tasks in accuracy, ITR and illiteracy/inefficiency, reaching an average ITR** of 52,8 bits/min and a maximum ITR** of 104,2 bits/min. Additionally, all participants achieved an accuracy level above 70% (illiteracy/inefficiency threshold) in both SSVEP and P300 tasks. Furthermore, the average accuracy of all tasks did not deteriorate if a reduced set with only the 8 best performing electrodes were used. These results are relevant for the development of online BMIs, including aspects related to usability, user satisfaction and portability. / A interface cérebro-máquina (ICM) é um sistema que permite a comunicação entre o sistema nervoso central e um dispositivo externo (Wolpaw et al., 2002). Aplicações de ICMs incluem o controle de próteses externa, cursores e teclados virtuais, para citar alguns. As ICMs desenvolvidas por vários grupos de pesquisa diferem em suas características (por exemplo, contínua ou discreta, síncrona ou assíncrona, graus de liberdade, outras) e, apesar de várias iniciativas voltadas para diretrizes de padronização, a comparação entre os estudos continua desafiadora (Brunner et al. 2015, Thompson et al., 2014). Aqui, utilizamos um equipamento EEG de 64 canais para adquirir dados de 19 participantes saudáveis ao longo da execução de três diferentes tarefas (SSVEP, P300 e híbrida) que permitiram quatro escolhas ao usuário e não exigiram nenhum treinamento prévio. Comparamos sistematicamente o desempenho \"off-line\" das três tarefas nos seguintes parâmetros: a) acurácia, b) taxa de transferência de informação, c) analfabetismo / ineficiência e d) preferências individuais. Além disso, selecionamos os melhores canais por tarefa e avaliamos a acurácia em função do número de eletrodos. Nossos resultados demonstraram que a tarefa SSVEP superou as demais em acurácia, ITR e analfabetismo/ineficiência, atingindo um ITR** médio de 52,8 bits/min e um ITR** máximo de 104,2 bits/min. Adicionalmente, todos os participantes alcançaram um nível de acurácia acima de 70% (limiar de analfabetismo/ineficiência) nas tarefas SSVEP e P300. Além disso, a acurácia média de todas as tarefas não se deteriorou ao se utilizar um conjunto reduzido composto apenas pelos melhores 8 eletrodos. Estes resultados são relevantes para o desenvolvimento de ICMs \"online\", incluindo aspectos relacionados à usabilidade, satisfação do usuário e portabilidade.

A brain-machine interface (BMI) is a system that allows the communication between the central nervous system (CNS) and an external device (Wolpaw et al. 2002). Applications of BMIs include the control of external prostheses, cursors and spellers, to name a few. The BMIs developed by various research groups differ in their characteristics (e.g. continuous or discrete, synchronous or asynchronous, degrees of freedom, others) and, in spite of several initiatives towards standardization and guidelines, the cross comparison across studies remains a challenge (Brunner et al. 2015; Thompson et al. 2014). Here, we used a 64-channel EEG equipment to acquire data from 19 healthy participants during three different tasks (SSVEP, P300 and hybrid) that allowed four choices to the user and required no previous neurofeedback training. We systematically compared the offline performance of the three tasks on the following parameters: a) accuracy, b) information transfer rate, c) illiteracy/inefficiency, and d) individual preferences. Additionally, we selected the best performing channels per task and evaluated the accuracy as a function of the number of electrodes. Our results demonstrate that the SSVEP task outperforms the other tasks in accuracy, ITR and illiteracy/inefficiency, reaching an average ITR** of 52,8 bits/min and a maximum ITR** of 104,2 bits/min. Additionally, all participants achieved an accuracy level above 70% (illiteracy/inefficiency threshold) in both SSVEP and P300 tasks. Furthermore, the average accuracy of all tasks did not deteriorate if a reduced set with only the 8 best performing electrodes were used. These results are relevant for the development of online BMIs, including aspects related to usability, user satisfaction and portability. / A interface cérebro-máquina (ICM) é um sistema que permite a comunicação entre o sistema nervoso central e um dispositivo externo (Wolpaw et al., 2002). Aplicações de ICMs incluem o controle de próteses externa, cursores e teclados virtuais, para citar alguns. As ICMs desenvolvidas por vários grupos de pesquisa diferem em suas características (por exemplo, contínua ou discreta, síncrona ou assíncrona, graus de liberdade, outras) e, apesar de várias iniciativas voltadas para diretrizes de padronização, a comparação entre os estudos continua desafiadora (Brunner et al. 2015, Thompson et al., 2014). Aqui, utilizamos um equipamento EEG de 64 canais para adquirir dados de 19 participantes saudáveis ao longo da execução de três diferentes tarefas (SSVEP, P300 e híbrida) que permitiram quatro escolhas ao usuário e não exigiram nenhum treinamento prévio. Comparamos sistematicamente o desempenho \"off-line\" das três tarefas nos seguintes parâmetros: a) acurácia, b) taxa de transferência de informação, c) analfabetismo / ineficiência e d) preferências individuais. Além disso, selecionamos os melhores canais por tarefa e avaliamos a acurácia em função do número de eletrodos. Nossos resultados demonstraram que a tarefa SSVEP superou as demais em acurácia, ITR e analfabetismo/ineficiência, atingindo um ITR** médio de 52,8 bits/min e um ITR** máximo de 104,2 bits/min. Adicionalmente, todos os participantes alcançaram um nível de acurácia acima de 70% (limiar de analfabetismo/ineficiência) nas tarefas SSVEP e P300. Além disso, a acurácia média de todas as tarefas não se deteriorou ao se utilizar um conjunto reduzido composto apenas pelos melhores 8 eletrodos. Estes resultados são relevantes para o desenvolvimento de ICMs \"online\", incluindo aspectos relacionados à usabilidade, satisfação do usuário e portabilidade.

Brain-computer interfaces (BCIs) blend methods and concepts researched by cognitive neuroscience, electrophysiology, computer science and engineering, resulting in systems of bi-directional information exchange directly between brain and computer. BCIs contribute to medical applications that restore communication and mobility for disabled patients and provide new forms of sending information to devices for enhancement and entertainment. Virtual reality (VR) introduces humans into a computer-generated world, tackling immersion and involvement. VR technology extends the classical multimedia experience, as the user is able to move within the environment, interact with other virtual participants, and manipulate objects, in order to generate the feeling of presence. This essay presents the possibilities of merging BCI with VR and the challenges to be tackled in the future. Current attempts to combine BCI and VR technology have shown that VR is a useful tool to test the functioning of BCIs, with safe, controlled and realistic experiments; there are better outcomes for VR and BCI combinations used for medical purposes compared to solely BCI training; and, enhancement systems for healthy users seem promising with VR-BCIs designed for home users. Future trends include brain-to-brain communication, sharing of several users’ brain signals within the virtual environment, and better and more efficient interfaces.

No description available.

Objective: Brain-computer interfaces (BCIs) aimed at restoring communication to people with severe neuromuscular disabilities often use event-related potentials (ERPs) in scalp-recorded EEG activity. Up to the present, most research and development in this area has been done in the laboratory with young healthy control subjects. In order to facilitate the development of BCI most useful to people with disabilities, the present study set out to: (1) determine whether people with amyotrophic lateral sclerosis (ALS) and healthy, age-matched volunteers (HVs) differ in the speed and accuracy of their ERP-based BCI use; (2) compare the ERP characteristics of these two groups; and (3) identify ERP-related factors that might enable improvement in BCI performance for people with disabilities. Methods: Sixteen EEG channels were recorded while people with ALS or healthy age-matched volunteers (HVs) used a P300-based BCI. The subjects with ALS had little or no remaining useful motor control (mean ALS Functional Rating Scale-Revised 9.4 (±9.5SD) (range 0-25)). Each subject attended to a target item as the items in a 6. ×. 6 visual matrix flashed. The BCI used a stepwise linear discriminant function (SWLDA) to determine the item the user wished to select (i.e., the target item). Offline analyses assessed the latencies, amplitudes, and locations of ERPs to the target and non-target items for people with ALS and age-matched control subjects. Results: BCI accuracy and communication rate did not differ significantly between ALS users and HVs. Although ERP morphology was similar for the two groups, their target ERPs differed significantly in the location and amplitude of the late positivity (P300), the amplitude of the early negativity (N200), and the latency of the late negativity (LN). Conclusions: The differences in target ERP components between people with ALS and age-matched HVs are consistent with the growing recognition that ALS may affect cortical function. The development of BCIs for use by this population may begin with studies in HVs but also needs to include studies in people with ALS. Their differences in ERP components may affect the selection of electrode montages, and might also affect the selection of presentation parameters (e.g., matrix design, stimulation rate). Significance: P300-based BCI performance in people severely disabled by ALS is similar to that of age-matched control subjects. At the same time, their ERP components differ to some degree from those of controls. Attention to these differences could contribute to the development of BCIs useful to those with ALS and possibly to others with severe neuromuscular disabilities.

In this work, the problem of detecting neuronal spikes or action potentials (AP) in noisy recordings from a Microelectrode Array (MEA) is investigated. In particular, the spike detection algorithms should be less complex and with low computational complexity so as to be amenable for real time applications. The use of the MEA is that it allows collection of extracellular signals from either a single unit or multiple (45) units within a small area. The noisy MEA recordings then undergo basic filtering, digitization and are presented to a computer for further processing. The challenge lies in using this data for detection of spikes from neuronal firings and extracting spatiotemporal patterns from the spike train which may allow control of a robotic limb or other neuroprosthetic device directly from the brain. The aim is to understand the spiking action of the neurons, and use this knowledge to devise efficient algorithms for Brain Machine Interfaces (BMIs). An effective BMI will require a realtime, computationally efficient implementation which can be carried out on a DSP board or FPGA system. The aim is to devise algorithms which can detect spikes and underlying spatio-temporal correlations having computational and time complexities to make a real time implementation feasible on a specialized DSP chip or an FPGA device. The time-frequency localization, multiresolution representation and analysis properties of wavelets make them suitable for analysing sharp transients and spikes in signals and distinguish them from noise resembling a transient or the spike. Three algorithms for the detection of spikes in low SNR MEA neuronal recordings are proposed: 1. A wavelet denoising method based on the Discrete Wavelet Transform (DWT) to suppress the noise power in the MEA signal or improve the SNR followed by standard thresholding techniques to detect the spikes from the denoised signal. 2. Directly thresholding the coefficients of the Stationary (Undecimated) Wavelet Transform (SWT) to detect the spikes. 3. Thresholding the output of a Teager Energy Operator (TEO) applied to the signal on the discrete wavelet decomposed signal resulting in a multiresolution TEO framework. The performance of the proposed three wavelet based algorithms in terms of the accuracy of spike detection, percentage of false positives and the computational complexity for different types of wavelet families in the presence of colored AR(5) (autoregressive model with order 5) and additive white Gaussian noise (AWGN) is evaluated. The performance is further evaluated for the wavelet family chosen under different levels of SNR in the presence of the colored AR(5) and AWGN noise. Chapter 1 gives an introduction to the concept behind Brain Machine Interfaces (BMIs), an overview of their history, the current state-of-the-art and the trends for the future. It also describes the working of the Microelectrode Arrays (MEAs). The generation of a spike in a neuron, the proposed mechanism behind it and its modeling as an electrical circuit based on the Hodgkin-Huxley model is described. An overview of some of the algorithms that have been suggested for spike detection purposes whether in MEA recordings or Electroencephalographic (EEG) signals is given. Chapter 2 describes in brief the underlying ideas that lead us to the Wavelet Transform paradigm. An introduction to the Fourier Transform, the Short Time Fourier Transform (STFT) and the Time-Frequency Uncertainty Principle is provided. This is followed by a brief description of the Continuous Wavelet Transform and the Multiresolution Analysis (MRA) property of wavelets. The Discrete Wavelet Transform (DWT) and its filter bank implementation are described next. It is proposed to apply the wavelet denoising algorithm pioneered by Donoho, to first denoise the MEA recordings followed by standard thresholding technique for spike detection. Chapter 3 deals with the use of the Stationary or Undecimated Wavelet Transform (SWT) for spike detection. It brings out the differences between the DWT and the SWT. A brief discussion of the analysis of non-stationary time series using the SWT is presented. An algorithm for spike detection based on directly thresholding the SWT coefficients without any need for reconstructing the denoised signal followed by thresholding technique as in the first method is presented. In chapter 4 a spike detection method based on multiresolution Teager Energy Operator is discussed. The Teager Energy Operator (TEO) picks up localized spikes in signal energy and thus is directly used for spike detection in many applications including R wave detection in ECG and various (alpha, beta) rhythms in EEG. Some basic properties of the TEO are discussed followed by the need for a multiresolution approach to TEO and the methods existing in literature. The wavelet decomposition and the subsampled signal involved at each level naturally lends it to a multiresolution TEO framework at the same time significantly reducing the computational complexity due the subsampled signal at each level. A wavelet-TEO algorithm for spike detection with similar accuracies as the previous two algorithms is proposed. The method proposed here differs significantly from that in literature since wavelets are used instead of time domain processing. Chapter 5 describes the method of evaluation of the three algorithms proposed in the previous chapters. The spike templates are obtained from MEA recordings, resampled and normalized for use in spike trains simulated as Poisson processes. The noise is modeled as colored autoregressive (AR) of order 5, i.e AR(5), as well as Additive White Gaussian Noise (AWGN). The noise in most human and animal MEA recordings conforms to the autoregressive model with orders of around 5. The AWGN Noise model is used in most spike detection methods in the literature. The performance of the proposed three wavelet based algorithms is measured in terms of the accuracy of spike detection, percentage of false positives and the computational complexity for different types of wavelet families. The optimal wavelet for this purpose is then chosen from the wavelet family which gives the best results. Also, optimal levels of decomposition and threshold factors are chosen while maintaining a balance between accuracy and false positives. The algorithms are then tested for performance under different levels of SNR with the noise modeled as AR(5) or AWGN. The proposed wavelet based algorithms exhibit a detection accuracy of approximately 90% at a low SNR of 2.35 dB with the false positives below 5%. This constitutes a significant improvement over the results in existing literature which claim an accuracy of 80% with false positives of nearly 10%. As the SNR increases, the detection accuracy increases to close to 100% and the false alarm rate falls to 0. Chapter 6 summarizes the work. A comparison is made between the three proposed algorithms in terms of detection accuracy and false positives. Directions in which future work may be carried out are suggested.