Banca de DEFESA: PAULA GABRIELLY RODRIGUES

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
STUDENT : PAULA GABRIELLY RODRIGUES
DATE: 12/12/2023
TIME: 14:00
LOCAL: https://conferenciaweb.rnp.br/webconf/diogo-10
TITLE:

Dynamic Brain Functional Connectivity in Electroencephalographic Signals Applied to the Development of Motor Imagery Brain-Computer Interfaces

PAGES: 200
BIG AREA: Engenharias
AREA: Engenharia Biomédica
SUBÁREA: Bioengenharia
SPECIALTY: Processamento de Sinais Biológicos
SUMMARY:

Graph theory has been very successful in efficiently representing the relationship between objects or elements of different natures, and proposals involving its use in characterizing brain functioning have proven to be an interesting option in developing brain-computer interfaces. (BCIs), as they allow to characterize the brain based on electroencephalography (EEG) signals acquired during different mental tasks, such as motor imagery (MI). The quantification and assessment of functional connectivity have been applied to studying brain organization during motor tasks to seek a better understanding of the imagery process that allows the development of more robust BCIs. However, until now, the dynamics of the imagery process and the interaction between different brain areas over time have rarely been considered. Recent studies suggest that these dynamics may provide relevant information for the discrimination between movement and rest, as well as in distinguishing between different MI tasks. With this in mind, the objective of this work is to evaluate the applicability of dynamic functional connectivity in these two scenarios, rest vs. MI and right-hand imagery vs. left-hand, seeking to extract relevant information for distinguishing different mental states. For this, three similarity measures and four graph metrics were estimated considering the static and dynamic approaches to compare and better characterize the dynamic connectivity. Furthermore, the most relevant electrodes for differentiating mental states were analyzed. In the case of the rest vs. MI, the dynamic connectivity allowed the distinction of states over time. To study the differentiation of MI tasks, a strategy based on a linear classifier and a complementary analysis based on deep learning (DL) were used due to limitations observed in the linear case. The results obtained point to the possible use of dynamic functional connectivity in the classification of mental tasks in MI-BCIs, but it did not prove to be advantageous in the context of DL strategies compared to what is commonly done in the literature.

COMMITTEE MEMBERS:
Presidente - Interno ao Programa - 1946319 - DIOGO COUTINHO SORIANO
Membro Titular - Examinador(a) Interno ao Programa - 1761107 - RICARDO SUYAMA
Membro Titular - Examinador(a) Interno ao Programa - 1544392 - ALINE DE OLIVEIRA NEVES PANAZIO
Membro Titular - Examinador(a) Externo à Instituição - THIAGO BULHOES DA SILVA COSTA - UNIFESP
Membro Titular - Examinador(a) Externo à Instituição - ROMIS RIBEIRO DE FAISSOL ATTUX - UNICAMP
Membro Suplente - Examinador(a) Interno ao Programa - 1672975 - JOAO RICARDO SATO
Membro Suplente - Examinador(a) Interno ao Programa - 2334927 - ANDRE KAZUO TAKAHATA
Membro Suplente - Examinador(a) Externo ao Programa - 1955999 - ANDRE MASCIOLI CRAVO
Membro Suplente - Examinador(a) Externo à Instituição - SARAH NEGREIROS DE CARVALHO LEITE - ITA