Mestrado em Bioética / Master Degree Course in Bioethics / Com este trabalho procuramos analisar a problemática das novas tecnologias no acesso ao conteúdo do cérebro, nomeadamente o uso dos potenciais evocados cognitivos (sobretudo o P300) sob a forma de Brain Fingerprinting, pondo à prova a fiabilidade do método e eventuais consequências éticas da sua aplicação. Abordaram-se conceitos como mente e seu conteúdo, no que se refere à comunicação e linguagem como forma de interacção social, sobretudo na capacidade de fazer uso da verdade ou da mentira, como escolha consciente e individual. Inerente à mentira abordou-se o conceito de emoção para compreender como funciona o tradicional polígrafo e de que forma o Brain Fingerprinting poderá trazer vantagens. A nível experimental avaliamos a utilização do P300 como "detector de culpa" em 20 indivíduos, tendo-se excluído 5, divididos num grupo controlo (n=8) e experimental (n=7 participantes num furto simulado), em duas experiências com estímulos visuais: experiência 1 com palavras e experiência 2 com imagens. Na análise visual individual verificou-se maior eficácia das imagens para a detecção correcta dos indivíduos (50% vs. 37,5% para os inocentes e 28,6% vs. 0% para os culpados). Nos culpados as palavras deram origem a uma percentagem elevada de falsos negativos (57,1% vs. 28,6% com imagens), salientando-se uma baixa taxa de falsos positivos em ambas as experiências. Na análise visual dos grupos as imagens parecem ser também mais eficazes na detecção correcta de ambos os perfis (culpados e inocentes), enquanto as palavras só identificam correctamente os inocentes. Da análise estatística intra e intergrupos obteve-se um "perfil inocente" em ambos os grupos, nas duas experiências. Neste estudo os nossos melhores resultados só identificaram correctamente 50% dos inocentes e 28,6% dos culpados (com imagens) o que se encontra muito longe dos 100% obtidos por Farwell e Smith (2001). Por fim, levamos a cabo uma discussão bioética debruçando-nos sobre a licitude deste tipo de metodologia, que promete aceder aos pensamentos, criando uma nova noção de dignidade, privacidade e autonomia. A discussão termina com a noção de responsabilização individual, abrangendo o direito inalienável à liberdade de acção e pensamento, inerente à espécie humana. / With the present work we intend to approach the problem of the use of new technologies as a mean to gain access to mental events, such as the use of Brain Fingerprinting. We will test the reliability of such method and discuss its ethical implications. We will approach concepts such as the mind and its content, concerning communication and language as a social interaction tool, and the ability to tell the truth or lie as a conscious and individual choice. Concerning deception, we approach the concept of emotion in order to understand the functioning of the traditional Polygraph and the way in which the use of Brain Fingerprinting may be advantageous. We experimentally assessed the use of the P300 event related potential as a "guilt detector", relying on a sample of 20 subjects (15+5 excluded from the analysis), divided into control (n = 8) and experimental (n = 7 subjects who simulated a theft in a mock-crime scenario) groups. There were two experiments with visual stimuli: experiment 1 with words and experiment 2 with pictures. Visual analysis of individual records revealed greater accuracy on the use of images to correctly detect subjects (50% vs. 37,5% for innocents and 28,6% vs. 0% of correctly detected subjects for the guilty condition). For guilty subjects, words led to a high percentage of false negatives (57,1% vs. 28,6 for pictures). A low percentage of false alarms were observed for both experiences. From the visual analysis of groups we are led to conclude that pictures seem to be once again more effective on the detection of both profiles (guilty and innocent), while words are only effective for the detection of innocent subjects. Statistical analysis did not discriminate between profiles, as we obtained constant "innocent" results for both experiments and groups. On the present study the best results we achieved were a 50% rate of identification for innocents and 28,6% for guilty subjects (with pictures). This contrasts with the rate of 100% obtained by Farwell and Smith (2001). Finally, we entail on a bioethical discussion concerning the acceptability of this kind of methodology, which promises access to thoughts, thus creating a new notion of dignity, privacy and autonomy. The discussion ends with the notion of individual responsibility, embracing the inalienable right to freedom of action and thought, inherent to the human species.

The use of electroencephalogram (EEG), an electrophysiological monitoring method for recording the brain activity, for authentication has attracted the interest of researchers for over a decade. In addition to exhibiting qualities of biometric-based authentication, they are revocable, impossible to mimic, and resistant to coercion attacks. However, EEG signals carry a wealth of information about an individual and can reveal private information about the user. This brings significant privacy issues to EEG-based authentication systems as they have access to raw EEG signals. This thesis proposes a privacy-preserving EEG-based authentication system that preserves the privacy of the user by not revealing the raw EEG signals while allowing the system to authenticate the user accurately. In that, perceptual hashing is utilized and instead of raw EEG signals, their perceptually hashed values are used in the authentication process. In addition to describing the authentication process, algorithms to compute the perceptual hash are developed based on two feature extraction techniques. Experimental results show that an authentication system using perceptual hashing can achieve performance comparable to a system that has access to raw EEG signals if enough EEG channels are used in the process. This thesis also presents a security analysis to show that perceptual hashing can prevent information leakage.