Detalhes bibliográficos
| Ano de defesa: |
2019 |
| Autor(a) principal: |
SANTOS, Alex Martins
 |
| Orientador(a): |
PAIVA, Anselmo Cardoso de
|
| Banca de defesa: |
PAIVA, Anselmo Cardoso de
,
SILVA, Aristófanes Corrêa
,
CONCI, Aura
,
VERAS, Rodrigo de Melo Souza
,
SANTANA, Ewaldo Eder Carvalho
|
| Tipo de documento: |
Tese
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal do Maranhão
|
| Programa de Pós-Graduação: |
PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA DE ELETRICIDADE/CCET
|
| Departamento: |
DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Palavras-chave em Inglês: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
https://tedebc.ufma.br/jspui/handle/tede/2744
|
Resumo: |
In this thesis, we propose a fully automated method for age-related macular degeneration (AMD) diagnosis based on image processing and computer learning techniques applied to OCT. AMD is a degenerative retinal disease considered to be one of the leading causes of blindness in the world’s elderly population. OCT is one of the major imaging tests for the detection and the monitoring of AMD, being a high-resolution biological imaging technology that allows the three-dimensional visualization of the internal structures of the eye on a micrometer scale. The evaluation of the generated images comes about through the specialist’s evaluation of the successive image slices in the search for morphological alterations in the macular region. Thus, the main objective of this thesis is the development and validation of an automatic detection of AMD, based on OCT images, which has high sensitivity and specificity for disease detection, reducing the need for large quantity visualization by the specialist for screening and detection. The main issues addressed in this work are the segmentation of the total retina and retinal pigment epithelial layer, as well as the methodology for AMD diagnosis. The proposed segmentation method is based on the use of deep-learning network CapsNet and graph cutting technique. The diagnostic method is based on geostatistical descriptors calculated on topographic maps of the retina. The classification uses support vector machine. The performed tests with 76 volumes of OCT reached an area under the ROC curve value of 0.996. |
