Detalhes bibliográficos
| Ano de defesa: |
2016 |
| Autor(a) principal: |
Much, Maicon Diogo
 |
| Orientador(a): |
Franco, Alexandre Rosa
|
| Banca de defesa: |
Não Informado pela instituição |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Pontifícia Universidade Católica do Rio Grande do Sul
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Engenharia Elétrica
|
| Departamento: |
Faculdade de Engenharia
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
http://tede2.pucrs.br/tede2/handle/tede/6835
|
Resumo: |
Patient head motion can cause failure in the analysis of functional magnetic resonance imaging (fMRI). Even very small movements (less than 5 mm) can corrupt volumes (3- dimensional images), invalidating the assumption that the variation in signal intensity between volumes is exclusive of changes in brain physiology. This behavior can cause problems such as: changes in neural activation patterns, emergence of false regions of activity and on many occasions the exam must be discarded. In an attempt to reduce the number of exams excluded due to excessive head motion, we developed a system capable of detecting individuals who have excessive movement while the images are being acquired. This real time analysis allows the test to be stopped and the subject can be redirected to not move. The implementation of this system was used based on the functional magnetic resonance imaging project in Real Time (rtFMRI) of AFNI platform. The system performance was evaluated in a simulated environment using an image database containing 162 acquisitions of functional images later collected from 81 volunteers. In addition, the system was also evaluated in real time during acquisition of RMF 10 tests. In a simulated environment, the motion detection method proposed in this project demonstrated an ability to detect early a failure in the exams in more than 75% of the cases. |
