Caracterização de sistemas ictogênicos perante extração de parâmetros eletrográficos

• Ano de defesa: 2015
• Autor(a) principal: Vinicius Rezende Carvalho
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Engenharia elétrica; Epilepsia; Processamento de sinais
• Link de acesso: http://hdl.handle.net/1843/RAOA-BC6G2K

This work deals with the extraction of features related to excitability and synchronism in electrographic signals, for evaluating hypotheses about auditory processing in WAR animals, an animal model for audiogenic seizures. These hypotheses are evaluated in Local Field Potential (LFP) signals from the Inferior Colliculus, obtained through experiment protocols based on the acoustic probing of the brain. Amplitude Modulated tones are presented to evoke Auditory Steady State Responses (ASSR). These can be considered as electrographic signatures of the auditory processing over time of each animal, highlighting features related to response magnitude and synchronism. The first protocol employs tones with intensity of 85 dB, enough to evoke the ASSR, but without triggering audiogenic seizures, modulated in amplitude at 53.71 Hz or 92.77 Hz, in interleaved trials. The first hypothesis states that the auditory processing in WARs is distinct from physiologic animals. The application of auditory stimuli would be needed to elicit responses which would make this difference visible. The larger amplitude response in WARs confirmed the hypothesis, indicating the hyperexcitability of the Inferior Colliculus of these animals. However, basal activity is also different between the groups, with ictogenic animals having greater energy in high frequency bands (gamma and beta), besides the occurrence of episodes resembling High Amplitude Spindles. The second hypothesis proposed that neural circuits in WARs have a tendency to synchronize more easily with external stimuli. This was confirmed by the higher phase coupling index (PLV) between stimulus and response of these animals. A diagnostic tool is then proposed, on which these features extracted from the probing protocol are used as inputs to classifiers that identify animals prone to audiogenic seizures. The ASSR was found to have the greatest predictive power, achieving accuracy of 96% with a modulating frequency of 53 Hz. The use of other features such as energies of endogenous EEG rhythms enables the improvement of performance, when a modulating frequency of 92 Hz is used. This indicates the possibility of using features from a probing protocol as diagnostic biomarkers. The second protocol involves the presentation of high intensity (110dB) AM tones in order to elicit audiogenic seizures in WARs only. The dynamic of the features presented earlier is analyzed in the pre-ictal, ictal and post-ictal periods, with the purpose of evaluating the hypothesis that the auditory processing in WARs is impaired during and after a seizure. It was shown that, although transitory changes happen during the seizure, the energy of the evoked response is not significantly changed across these three periods. This contrasts with the energies in other frequency bands, which significantly increase during the seizure due to recruitment by the ictal focus, followed by a refractory period with decreased activity after the seizure. The synchrony of the response with the stimulus, measured by the PLV index, confirms part of this conclusion. Before the seizure, the PLV is below the normal range found in WARs, possibly due to interactions of the response circuits with ictogenic ones. Thus, despite the magnitude of the response not being changed before the seizure, the ability to follow the stimulus is slightly impaired. During the seizure, there are only transient changes of the PLV, especially near its end. In the post-ictal period, most of the endogenous oscillations are in refractory period, but the system is still responsive to external sound stimuli, as evidenced by the persistence and synchrony of the ASSR. Thus, the dynamics of the ASSR suggest that the circuitry of the auditory xi pathway, despite being involved in the processes leading to audiogenic seizures, may not effectively be part of the ictogenic focus.