Análisis de señales EEG con potencial uso en el control de una mano robótica por medio de la aplicación de imaginación motora

Abstract

This work as a degree project is based on the analysis of electroencephalography (EEG) signals under the application of motor imagination (MI), specifically on the intention of hand movement, thus generating a brain-computer interface. The signals to be studied were downloaded from the free database EEG Motor Movement / Imagery Dataset (eegmmidb) of PHYSIONET, these were acquired through the BCI2000 system where the electrodes are located superficially in healthy patients. The extraction of characteristics to apply the mental strategy (motor imagination), will be carried out through the application of the discrete Wavelet transform (daubechies) and for its classification two models are used based on: vector support machines and a feedforward neural network, visualizing the precision of each of the models used in the way of successes and failures through a confusion matrix