EOG/EEG ACQUISITION AND ANALYSIS FOR DISCRIMINATION OF TYPICAL RESPONSES IN THE HIGH PASS BAND

• Autor(a) principal: Roxo, Mariana Garcês Meira
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/160609
• topic: EEG; EOG; m-sequences; matched filter; saccades; Domínio/Área Científica::Engenharia e Tecnologia::Outras Engenharias e Tecnologias

The link between saccadic movements and neurological diseases has proven to be interesting, since the former change as a result of the latter. These diseases are often challenging to diagnose, as they may already be at an extremely developed stage at the time of diagnosis. In this thesis, these movements were used in order to develop a model of the transmission of information in the brain, aiming at investigating typical response patterns in detection of the transmitted information. For this purpose, 6 subjects were presented with a slide show, designed using a 127 msequence, as to avoid any learning phenomenon. During the experiment, electroencephalography (EEG) and electrooculography (EOG) signals were collected. An algorithm was then developed whose goal was to estimate the previously presented sequence using only the signals collected above certain frequencies. Subsequently, typical responses in detection were analyzed. For all subjects, only one sequence was correctly detected, namely the one that had been selected to be shown. With increasing cutoff frequency, the number of detections tended to increase. At lower cutoff frequencies, the number of detections was substantially lower for one of the subjects. For three subjects, rates of 100% were reached, which were considered abnormal. In summary, the algorithm proved to be efficient in estimating the sequences using the EEG and EOG signals as objects of analysis. In the future, if the algorithm is tested on subjects with pathology, it is proposed that healthy subjects will show non-pathological patterns and unhealthy subjects will show patterns of pathological ones. If this hypothesis is confirmed, this algorithm could contribute to a potential predictor of a biomarker for these diseases in the future.