Classify four imagined objects with EEG signals

EEG signals contain information directly related to cognitive activity. This paper presents a method classify the images person imagines via provided by signals. The relating objects ‘tree’, ‘house’, ‘plane’ and ‘dog’ have been reconstructed. We used convolutional neural networks obtain reconstruction of genetic algorithm find parameters this network. results obtained evaluated means Chebychev metric compare images, it shows that is performed with success 57% over chance, an accuracy in classification 60% kappa value 0.40, demonstrating five mental states where four them come from visual imagery, possible.