Estudio de la valencia afectiva de estímulos visuales mediante técnicas de clasificación de señales EEG

Resumen

New approaches on the study of emotions to obtain objective measures from different affective states create a great interest for neuroscientists, in particular, due to the potential applications of new methods on the analysis of neuropathologies and psychological disorders. In the present thesis project, the development of pattern recognition algorithms, i. e. classifiers, has proven to be very helpful for extracting relevant features from neurophysiological signals. On one hand, the application of these techniques has showed to be more sensitive to find relevant differences between two experimental conditions (positive and negative affective valence) than other conventional analyses. Furthermore, the performed analyses have showed that some determined spectral rhythms, topographical locations and time intervals are specially linked to the emotional processing of visual stimuli. Moreover, the application of pattern recognition algorithms, for instance an algorithm based on support vectors machine, on a set of participants scoring both low and high in a neuroticism scale, has also proven the suitability of this technique in order to study the influence of the individual differences on the particular brain responses of human beings. In conclusion, the application of classification techniques on EEG signals suggests that these approaches are really useful to better understanding the psychobiological basis of the human behavior and can be taken into consideration on diverse neuropathological evaluations and emotional monitoring.