Genetic feature selection to optimally detect P300 in brain computer interfaces.

A Brain Computer Interface is a system that provides an artificial communication between the human brain and the external world. The paradigm based on event related evoked potentials is used in this work. Our main goal was to efficiently solve a binary classification problem: presence or absence of P300 in the registers. Genetic Algorithms and Support Vector Machines were used in a wrapper configuration for feature selection and classification. The original input patterns were provided by two channels (Oz and Fz) of resampled EEG registers and wavelet coefficients. To evaluate the performance of the system, accuracy, sensibility and specificity were calculated. The wrapped wavelet patterns show a better performance than the temporal ones. The results were similar for patterns from channel Oz and Fz, together or separated.

INTERFAZ CEREBRO COMPUTADORA (ICC) BASADA EN EL POTENCIAL RELACIONADO CON EVENTOS P300: ANÁLISIS DEL EFECTO DE LA DIMENSIÓN DE LA MATRIZ DE ESTIMULACIÓN SOBRE SU DESEMPEÑO

Una interfaz cerebro computadora (ICC) es un dispositivo que ayuda a personas con deficiencias motoras severas, al permitir la realización de una comunicación externa a partir de la actividad eléctrica del cerebro sin la asistencia de los nervios periféricos o de la actividad muscular, prometiendo además una mejora en la calidad de vida de los pacientes. En este proyecto se utilizó un sistema ICC basado en el paradigma P300, desarrollado en la Universidad Nacional de Entre Ríos. El sistema cuenta con un sistema no invasivo de adquisición de electroencefalograma, un amplificador Grass, el software BCI2000 y el paquete de simulación robótica Marilou. Adicionalmente, el sistema permite evaluar la aplicación de dicha ICC en el control de una silla de ruedas autopropulsada e inteligente. La presentación de estímulos para la generación del P300 se llevó a cabo con matrices de íconos que codifican las instrucciones de comandos o direcciones para la silla de ruedas. En el presente trabajo se probaron dos matrices con diferentes dimensiones y distribuciones, la primera de 4x5 y la segunda de 4x3. Se analizaron los porcentajes de clasificación que éstas arrojaron con el método de regresión SWLDA, donde se concluyó que la matriz de 4x3 presentaba mayores porcentajes de clasificación que la matriz 4x5. Las implicaciones con respecto al control de la silla se vislumbran como mayor confort y exactitud en el sistema inteligente.

A brain computer interface BCI is a device that helps people with severs motor disabilities. It allows an external communication through the electrical activity of the brain without the assistance of the peripheral nerves or muscle activity. This project used a BCI system, based on P300 paradigm which was developed at Universidad Nacional de Entre Ríos. The system includes an EEG signal acquisition system that use external electrodes, a Grass amplifier, the BCI2000 software, and the Marilou robotic simulation tool. Additionally, the system allows the evaluation of the BCI application to control the movement of an intelligent and self-propelled wheelchair. The presentation of icons, which codified the instructions to command the wheelchair movements, was developed, in order to generate the stimulus for P300 generation. Two matrix with different size and distribution (4x5 and 4x3, row x column) were tested. We analyzed the percentage of classification obtained after the application of the regression method SWLDA, and we found that the major classification percentage was achieved with the 4x3 matrix. This study reveals that this process could be faster and more confortable for the user. And finally the subject decisions will have more correlation between the results of the system and his real desire.

Single trial P300 detection based on the Empirical Mode Decomposition.

We present a new method for single trial detection of P300 evoked responses. The features used to classify are the coefficients of a least-squares fit of a single EEG epoch to the intrinsical mode functions of an empirical mode decomposition of the averaged event response from a P300 training set. Support vector machines with a linear kernel are used to classify the epochs and receiver operating characteristic analysis is used to evaluate our method's performance.