Resumen

Our brain produces electrical impulses that travel through neurons, these impulses are called brain waves, depending on the mental state of the subject in question, these waves are divided into 4 types: delta, theta, alpha and beta. This project seeks to determine the level of concentration of people from an Analog Front End (AFE). To carry out this study, the compilation of the Electroencephalography (EEG) database was carried out using the Texas Instruments ADS1299EEG-FE acquisition card with its respective software. Subsequently, four classification methods were implemented in the MATLAB software, four methods were implemented, including method 1, which uses Power Spectral Density Analysis (PSD) that classifies using a neural network and a relationship between powers of alpha and beta bands, method 2, Hilbert Huang's Method (HHT) that classifies with a neural network, method 3, Common Spatial Pattern Classification (CSP), and method 4, Discrete Wavelet Transformed (DWT). These methods allow the mental state of patients to be determined (concentrated or not concentrated) with accuracy of success: FFT 85%, HHT 74%, CSP 87%, DWT 70%, for database acquired with AFE. Finally, the method that gave the best results is selected for the purpose of the project.