Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D
Autor(a) principal: | |
---|---|
Data de Publicação: | 2021 |
Tipo de documento: | Tese |
Idioma: | por |
Título da fonte: | Biblioteca Digital de Teses e Dissertações da UFMA |
Texto Completo: | https://tedebc.ufma.br/jspui/handle/tede/tede/3676 |
Resumo: | Prostate cancer is the second most common cancer in men in the world. In Brazil, there are an estimated 65,840 new cases of prostate cancer for each year of the 2020-2022 triennium. The automatic segmentation of the prostate is an important factor to assist in the diagnosis and treatment of cancer, such as orientation of the biopsy procedure and radiotherapy. However, automatic segmentation is challenging due to the great variation in the anatomy of the prostate due to pathological changes, tissue similar to Organs adjacent organs and different image acquisition protocols. Therefore, this work proposes three computational methods based on superpixels for automatic segmentation of the prostate in 3D magnetic resonance (MR) images. All the proposed methods consider the following steps: 1) description of the materials, 2) prostate detection, 3) image enhancement, 4) prostate segmentation, 5) refinement of the segmentation, and 6) evaluation of the results. The differences between the proposed methods are found in the segmentation of the prostate with the subset of superpixels classification. The first proposed method presents a classification approach based on the deep learning technique Convolutional Neural Network (CNN) and the particle swarm optimization algorithm (PSO) to optimize the filters in the convolutional layers, the second proposed method describes a classification approach conventional based on texture descriptors, using phylogenetic indices, the eXtreme Gradient Boosting (XGBoost) algorithm and the PSO algorithm to optimize the XGBoost hyperparameters, and finally, the third proposed method details a hybrid classification approach based on the CNN technique, the XGBoost algorithm and the PSO algorithm to optimize the type of connection used in the convolutional layers. The proposed methods were evaluated on the databases Prostate 3T and PROMISE12 using the performance metrics Dice similarity coefficient, relative volume difference, volumetric similarity, average distance surface and Hausdorff distance. The results of the application of the first method showed 87.67%, 2.83%, 0.96, 0.89 mm, and 13.65 mm, respectively, in the corresponding values of the mentioned performance metrics. The second proposed method obtained 85.64%, 7.68%, 0.96, 1.22 mm, and 15.13 mm, respectively. Finally, the proposed third method reached 87.65%, 3.18%, 0.96, 0.88 mm, and 13.51 mm, respectively. It was found that the first and the third method showed similar results in the segmentation of the prostate, being superior to the results obtained in the second method. In addition, the third method had a lower standard deviation in the metrics and a higher rate of hit in the prostate superpixels than the other methods. The experimental results demonstrate the performance potential of the proposed methods compared to those recently published in the literature. |
id |
UFMA_c88399ca74105ba301e6d2f2bf7b8e65 |
---|---|
oai_identifier_str |
oai:tede2:tede/3676 |
network_acronym_str |
UFMA |
network_name_str |
Biblioteca Digital de Teses e Dissertações da UFMA |
repository_id_str |
2131 |
spelling |
SILVA, Aristófanes Corrêahttp://lattes.cnpq.br/2446301582459104PAIVA, Anselmo Cardoso dehttp://lattes.cnpq.br/6446831084215512PAIVA, Anselmo Cardoso dehttp://lattes.cnpq.br/6446831084215512AIRES, Kelson Rômulo Teixeirahttp://lattes.cnpq.br/0065931835203045ARAÚJO, Flávio Henrique Duarte dehttp://lattes.cnpq.br/9403364226017898CASAS, Vicente Leonardo Paucarhttp://lattes.cnpq.br/1155686983267102BARROS NETTO, Stelmo Magalhãeshttp://lattes.cnpq.br/9897454292052062http://lattes.cnpq.br/4300669916019534SILVA, Giovanni Lucca França da2022-06-13T17:25:37Z2021-03-16SILVA, Giovanni Lucca França da. Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D. 2021. 128 f. Tese (Programa de Pós-Graduação em Engenharia de Eletricidade/CCET) - Universidade Federal do Maranhão, São Luís, 2021.https://tedebc.ufma.br/jspui/handle/tede/tede/3676Prostate cancer is the second most common cancer in men in the world. In Brazil, there are an estimated 65,840 new cases of prostate cancer for each year of the 2020-2022 triennium. The automatic segmentation of the prostate is an important factor to assist in the diagnosis and treatment of cancer, such as orientation of the biopsy procedure and radiotherapy. However, automatic segmentation is challenging due to the great variation in the anatomy of the prostate due to pathological changes, tissue similar to Organs adjacent organs and different image acquisition protocols. Therefore, this work proposes three computational methods based on superpixels for automatic segmentation of the prostate in 3D magnetic resonance (MR) images. All the proposed methods consider the following steps: 1) description of the materials, 2) prostate detection, 3) image enhancement, 4) prostate segmentation, 5) refinement of the segmentation, and 6) evaluation of the results. The differences between the proposed methods are found in the segmentation of the prostate with the subset of superpixels classification. The first proposed method presents a classification approach based on the deep learning technique Convolutional Neural Network (CNN) and the particle swarm optimization algorithm (PSO) to optimize the filters in the convolutional layers, the second proposed method describes a classification approach conventional based on texture descriptors, using phylogenetic indices, the eXtreme Gradient Boosting (XGBoost) algorithm and the PSO algorithm to optimize the XGBoost hyperparameters, and finally, the third proposed method details a hybrid classification approach based on the CNN technique, the XGBoost algorithm and the PSO algorithm to optimize the type of connection used in the convolutional layers. The proposed methods were evaluated on the databases Prostate 3T and PROMISE12 using the performance metrics Dice similarity coefficient, relative volume difference, volumetric similarity, average distance surface and Hausdorff distance. The results of the application of the first method showed 87.67%, 2.83%, 0.96, 0.89 mm, and 13.65 mm, respectively, in the corresponding values of the mentioned performance metrics. The second proposed method obtained 85.64%, 7.68%, 0.96, 1.22 mm, and 15.13 mm, respectively. Finally, the proposed third method reached 87.65%, 3.18%, 0.96, 0.88 mm, and 13.51 mm, respectively. It was found that the first and the third method showed similar results in the segmentation of the prostate, being superior to the results obtained in the second method. In addition, the third method had a lower standard deviation in the metrics and a higher rate of hit in the prostate superpixels than the other methods. The experimental results demonstrate the performance potential of the proposed methods compared to those recently published in the literature.O câncer de próstata é o segundo câncer mais frequente em homens no mundo. No Brasil, estimam-se 65.840 casos novos de câncer de próstata para cada ano do triênio 2020-2022. A segmentação automática da próstata é um fator importante para auxiliar o diagnóstico e o tratamento do câncer, como orientação do procedimento de biópsia e a radioterapia. No entanto, a segmentação automática é desafiadora devido à grande variação na anatomia da próstata por conta das alterações patológicas, tecido semelhante aos órgãos adjacentes e diferentes protocolos de aquisição das imagens. Portanto, esta tese propõe três métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética (RM) 3D. Todos os métodos propostos consideram as seguintes etapas: 1) descrição dos materiais, 2) detecção da próstata, 3) melhoramento das imagens, 4) segmentação da próstata, 5) refinamento da segmentação, e a 6) avaliação dos resultados. As diferenças entre os métodos propostos encontram-se na etapa de segmentação da próstata com a subetapa de classificação dos superpixels. O primeiro método proposto apresenta uma abordagem de classificação baseada na t´técnica de aprendizagem profunda Convolutional Neural Network (CNN) e o algoritmo de otimização por enxame de partículas (PSO) para otimizar os filtros nas camadas convolucionais, o segundo m´etodo proposto descreve uma abordagem de classificação convencional baseada em descritores de textura, usando os ´índices filogenéticos, o algoritmo eXtreme Gradient Boosting (XGBoost) e o algoritmo PSO para otimizar os hiperparâmetros do XGBoost, e por fim, o terceiro método proposto detalha uma abordagem de classificação híbrida baseada na técnica CNN, o algoritmo XGBoost e o algoritmo PSO para otimizar o tipo de conexão utilizada nas camadas convolucionais. Os métodos propostos foram avaliados nas bases de imagens de RM 3D Prostate 3T e PROMISE12 usando as métricas de desempenho coeficiente de similaridade Dice, volume relativo da diferença, similaridade volumétrica, distância média da superfície e distância de Hausdorff. Os resultados da aplicação do primeiro método apresentaram 87,67%, 2,83%, 0,96, 0,89 mm, e 13,65 mm, respectivamente nos valores correspondentes das métricas de desempenho mencionadas. O segundo método proposto obteve 85,64%, 7,68%, 0,96, 1,22 mm, e 15,13 mm, respectivamente. Finalmente, o terceiro método proposto alcançou 87,65%, 3,18%, 0,96, 0,88 mm, e 13,51 mm, respectivamente. Constatou-se que o primeiro e o terceiro m´método apresentaram resultados similares na segmentação da próstata, sendo eles superiores aos resultados obtidos no segundo método. Além disso, o terceiro m´método apresentou um menor desvio padrão nas métricas e uma taxa de acerto no superpixels de próstata superior aos demais m´métodos. Os resultados experimentais demonstram o potencial de desempenho dos métodos propostos comparados aos publicados recentemente na literatura. Palavras-chave: Segmentação da próstata, Superpixels, Convolutional neural network, Índices filogenéticos, Algoritmo XGBoost, Otimização por enxame de partículasSubmitted by Daniella Santos (daniella.santos@ufma.br) on 2022-06-13T17:25:37Z No. of bitstreams: 1 GiovanniLuccaFrançadaSilva.pdf: 7192346 bytes, checksum: 71622678dd8a601b0721148844fb0e7f (MD5)Made available in DSpace on 2022-06-13T17:25:37Z (GMT). No. of bitstreams: 1 GiovanniLuccaFrançadaSilva.pdf: 7192346 bytes, checksum: 71622678dd8a601b0721148844fb0e7f (MD5) Previous issue date: 2021-03-16CAPESapplication/pdfporUniversidade Federal do MaranhãoPROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA DE ELETRICIDADE/CCETUFMABrasilDEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCETsegmentação da próstata;superpixels;convolutional neural network;índices filogenéticos;algoritmo xgboost;otimização por enxame de partículas;prostate segmentation,superpixels;convolutional neural networks;phylogenetic indexes;xGBoost algorithm;particle swarm optimizationAnálise de Algoritmos e Complexidade de ComputaçãoMétodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3DSuperpixel-based computational methods for automatic prostate segmentation in 3D magnetic resonance imaginginfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/doctoralThesisinfo:eu-repo/semantics/openAccessreponame:Biblioteca Digital de Teses e Dissertações da UFMAinstname:Universidade Federal do Maranhão (UFMA)instacron:UFMAORIGINALGiovanniLuccaFrançadaSilva.pdfGiovanniLuccaFrançadaSilva.pdfapplication/pdf7192346http://tedebc.ufma.br:8080/bitstream/tede/3676/2/GiovanniLuccaFran%C3%A7adaSilva.pdf71622678dd8a601b0721148844fb0e7fMD52LICENSElicense.txtlicense.txttext/plain; charset=utf-82255http://tedebc.ufma.br:8080/bitstream/tede/3676/1/license.txt97eeade1fce43278e63fe063657f8083MD51tede/36762022-06-13 14:25:37.992oai:tede2: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Biblioteca Digital de Teses e Dissertaçõeshttps://tedebc.ufma.br/jspui/PUBhttp://tedebc.ufma.br:8080/oai/requestrepositorio@ufma.br||repositorio@ufma.bropendoar:21312022-06-13T17:25:37Biblioteca Digital de Teses e Dissertações da UFMA - Universidade Federal do Maranhão (UFMA)false |
dc.title.por.fl_str_mv |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
dc.title.alternative.eng.fl_str_mv |
Superpixel-based computational methods for automatic prostate segmentation in 3D magnetic resonance imaging |
title |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
spellingShingle |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D SILVA, Giovanni Lucca França da segmentação da próstata; superpixels; convolutional neural network; índices filogenéticos; algoritmo xgboost; otimização por enxame de partículas; prostate segmentation, superpixels; convolutional neural networks; phylogenetic indexes; xGBoost algorithm; particle swarm optimization Análise de Algoritmos e Complexidade de Computação |
title_short |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
title_full |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
title_fullStr |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
title_full_unstemmed |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
title_sort |
Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D |
author |
SILVA, Giovanni Lucca França da |
author_facet |
SILVA, Giovanni Lucca França da |
author_role |
author |
dc.contributor.advisor1.fl_str_mv |
SILVA, Aristófanes Corrêa |
dc.contributor.advisor1Lattes.fl_str_mv |
http://lattes.cnpq.br/2446301582459104 |
dc.contributor.advisor-co1.fl_str_mv |
PAIVA, Anselmo Cardoso de |
dc.contributor.advisor-co1Lattes.fl_str_mv |
http://lattes.cnpq.br/6446831084215512 |
dc.contributor.referee1.fl_str_mv |
PAIVA, Anselmo Cardoso de |
dc.contributor.referee1Lattes.fl_str_mv |
http://lattes.cnpq.br/6446831084215512 |
dc.contributor.referee2.fl_str_mv |
AIRES, Kelson Rômulo Teixeira |
dc.contributor.referee2Lattes.fl_str_mv |
http://lattes.cnpq.br/0065931835203045 |
dc.contributor.referee3.fl_str_mv |
ARAÚJO, Flávio Henrique Duarte de |
dc.contributor.referee3Lattes.fl_str_mv |
http://lattes.cnpq.br/9403364226017898 |
dc.contributor.referee4.fl_str_mv |
CASAS, Vicente Leonardo Paucar |
dc.contributor.referee4Lattes.fl_str_mv |
http://lattes.cnpq.br/1155686983267102 |
dc.contributor.referee5.fl_str_mv |
BARROS NETTO, Stelmo Magalhães |
dc.contributor.referee5Lattes.fl_str_mv |
http://lattes.cnpq.br/9897454292052062 |
dc.contributor.authorLattes.fl_str_mv |
http://lattes.cnpq.br/4300669916019534 |
dc.contributor.author.fl_str_mv |
SILVA, Giovanni Lucca França da |
contributor_str_mv |
SILVA, Aristófanes Corrêa PAIVA, Anselmo Cardoso de PAIVA, Anselmo Cardoso de AIRES, Kelson Rômulo Teixeira ARAÚJO, Flávio Henrique Duarte de CASAS, Vicente Leonardo Paucar BARROS NETTO, Stelmo Magalhães |
dc.subject.por.fl_str_mv |
segmentação da próstata; superpixels; convolutional neural network; índices filogenéticos; algoritmo xgboost; otimização por enxame de partículas; |
topic |
segmentação da próstata; superpixels; convolutional neural network; índices filogenéticos; algoritmo xgboost; otimização por enxame de partículas; prostate segmentation, superpixels; convolutional neural networks; phylogenetic indexes; xGBoost algorithm; particle swarm optimization Análise de Algoritmos e Complexidade de Computação |
dc.subject.eng.fl_str_mv |
prostate segmentation, superpixels; convolutional neural networks; phylogenetic indexes; xGBoost algorithm; particle swarm optimization |
dc.subject.cnpq.fl_str_mv |
Análise de Algoritmos e Complexidade de Computação |
description |
Prostate cancer is the second most common cancer in men in the world. In Brazil, there are an estimated 65,840 new cases of prostate cancer for each year of the 2020-2022 triennium. The automatic segmentation of the prostate is an important factor to assist in the diagnosis and treatment of cancer, such as orientation of the biopsy procedure and radiotherapy. However, automatic segmentation is challenging due to the great variation in the anatomy of the prostate due to pathological changes, tissue similar to Organs adjacent organs and different image acquisition protocols. Therefore, this work proposes three computational methods based on superpixels for automatic segmentation of the prostate in 3D magnetic resonance (MR) images. All the proposed methods consider the following steps: 1) description of the materials, 2) prostate detection, 3) image enhancement, 4) prostate segmentation, 5) refinement of the segmentation, and 6) evaluation of the results. The differences between the proposed methods are found in the segmentation of the prostate with the subset of superpixels classification. The first proposed method presents a classification approach based on the deep learning technique Convolutional Neural Network (CNN) and the particle swarm optimization algorithm (PSO) to optimize the filters in the convolutional layers, the second proposed method describes a classification approach conventional based on texture descriptors, using phylogenetic indices, the eXtreme Gradient Boosting (XGBoost) algorithm and the PSO algorithm to optimize the XGBoost hyperparameters, and finally, the third proposed method details a hybrid classification approach based on the CNN technique, the XGBoost algorithm and the PSO algorithm to optimize the type of connection used in the convolutional layers. The proposed methods were evaluated on the databases Prostate 3T and PROMISE12 using the performance metrics Dice similarity coefficient, relative volume difference, volumetric similarity, average distance surface and Hausdorff distance. The results of the application of the first method showed 87.67%, 2.83%, 0.96, 0.89 mm, and 13.65 mm, respectively, in the corresponding values of the mentioned performance metrics. The second proposed method obtained 85.64%, 7.68%, 0.96, 1.22 mm, and 15.13 mm, respectively. Finally, the proposed third method reached 87.65%, 3.18%, 0.96, 0.88 mm, and 13.51 mm, respectively. It was found that the first and the third method showed similar results in the segmentation of the prostate, being superior to the results obtained in the second method. In addition, the third method had a lower standard deviation in the metrics and a higher rate of hit in the prostate superpixels than the other methods. The experimental results demonstrate the performance potential of the proposed methods compared to those recently published in the literature. |
publishDate |
2021 |
dc.date.issued.fl_str_mv |
2021-03-16 |
dc.date.accessioned.fl_str_mv |
2022-06-13T17:25:37Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/doctoralThesis |
format |
doctoralThesis |
status_str |
publishedVersion |
dc.identifier.citation.fl_str_mv |
SILVA, Giovanni Lucca França da. Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D. 2021. 128 f. Tese (Programa de Pós-Graduação em Engenharia de Eletricidade/CCET) - Universidade Federal do Maranhão, São Luís, 2021. |
dc.identifier.uri.fl_str_mv |
https://tedebc.ufma.br/jspui/handle/tede/tede/3676 |
identifier_str_mv |
SILVA, Giovanni Lucca França da. Métodos computacionais baseados em superpixels para segmentação automática da próstata em imagens de ressonância magnética 3D. 2021. 128 f. Tese (Programa de Pós-Graduação em Engenharia de Eletricidade/CCET) - Universidade Federal do Maranhão, São Luís, 2021. |
url |
https://tedebc.ufma.br/jspui/handle/tede/tede/3676 |
dc.language.iso.fl_str_mv |
por |
language |
por |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
Universidade Federal do Maranhão |
dc.publisher.program.fl_str_mv |
PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA DE ELETRICIDADE/CCET |
dc.publisher.initials.fl_str_mv |
UFMA |
dc.publisher.country.fl_str_mv |
Brasil |
dc.publisher.department.fl_str_mv |
DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET |
publisher.none.fl_str_mv |
Universidade Federal do Maranhão |
dc.source.none.fl_str_mv |
reponame:Biblioteca Digital de Teses e Dissertações da UFMA instname:Universidade Federal do Maranhão (UFMA) instacron:UFMA |
instname_str |
Universidade Federal do Maranhão (UFMA) |
instacron_str |
UFMA |
institution |
UFMA |
reponame_str |
Biblioteca Digital de Teses e Dissertações da UFMA |
collection |
Biblioteca Digital de Teses e Dissertações da UFMA |
bitstream.url.fl_str_mv |
http://tedebc.ufma.br:8080/bitstream/tede/3676/2/GiovanniLuccaFran%C3%A7adaSilva.pdf http://tedebc.ufma.br:8080/bitstream/tede/3676/1/license.txt |
bitstream.checksum.fl_str_mv |
71622678dd8a601b0721148844fb0e7f 97eeade1fce43278e63fe063657f8083 |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 |
repository.name.fl_str_mv |
Biblioteca Digital de Teses e Dissertações da UFMA - Universidade Federal do Maranhão (UFMA) |
repository.mail.fl_str_mv |
repositorio@ufma.br||repositorio@ufma.br |
_version_ |
1800303806623252480 |