Deep learning for melanoma classification : a study using skin lesion images
Autor(a) principal: | |
---|---|
Data de Publicação: | 2023 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10400.14/42247 |
Resumo: | Cutaneous melanoma is considered the skin cancer with highest mortality rate and has been gaining the attention of the medical community due to its rapidly increasing incidence. Advancements in computational technologies have paved the way for innovative image detection methods that can be transferable to medical ap plications, significantly enhancing the potential for early intervention in melanoma diagnosis. To make diagnosis more accurate and to further increase survival rates, this study employs deep learning techniques on an extensive dataset derived from multiple sources. Utilizing Microsoft Azure Cloud as the computational infras tructure, trial and error approach was employed by hyperparameterizing several convolutional neural networks (CNN) where the decision criteria were choosing the one with highest Fβ Score. MAR-MELA-CNN is an innovative ensemble model in corporating six state-of-the-art pre-trained CNN architectures: Xception, VGG16, ResNet50, NASNetMobile, MobileNetV2, and InceptionV3. The primary goal of this research is to further understand CNN’s efficiency in the diagnosis of melanoma and to furthermore measure its performance on a merged dataset. The proposed algorithm achieved a Fβ score of 85%, an area under the curve (AUC) score of 93%, and an average precision (AP) score of 92%, promising diagnostic tool for cutaneous melanoma compared to traditional methods. Further improvements lay in the improvement of the architecture, expansion of the computational instances as well as of the dataset. Another field of future work could be devising a strategy for real-time implementation of this model in a hospital setting, as it could be of vital importance to provide swift support to doctors. |
id |
RCAP_cb86c1dc59235d0e77380130b5881dbf |
---|---|
oai_identifier_str |
oai:repositorio.ucp.pt:10400.14/42247 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Deep learning for melanoma classification : a study using skin lesion imagesCutaneous melanomaDeep learningConvolutional neural networksFβ scoreMedical imagingMelanoma cutâneoAprendizagem profundaRedes neuronais convolucionaisPontuação FβImagiologia médicaDomínio/Área Científica::Ciências Sociais::Economia e GestãoCutaneous melanoma is considered the skin cancer with highest mortality rate and has been gaining the attention of the medical community due to its rapidly increasing incidence. Advancements in computational technologies have paved the way for innovative image detection methods that can be transferable to medical ap plications, significantly enhancing the potential for early intervention in melanoma diagnosis. To make diagnosis more accurate and to further increase survival rates, this study employs deep learning techniques on an extensive dataset derived from multiple sources. Utilizing Microsoft Azure Cloud as the computational infras tructure, trial and error approach was employed by hyperparameterizing several convolutional neural networks (CNN) where the decision criteria were choosing the one with highest Fβ Score. MAR-MELA-CNN is an innovative ensemble model in corporating six state-of-the-art pre-trained CNN architectures: Xception, VGG16, ResNet50, NASNetMobile, MobileNetV2, and InceptionV3. The primary goal of this research is to further understand CNN’s efficiency in the diagnosis of melanoma and to furthermore measure its performance on a merged dataset. The proposed algorithm achieved a Fβ score of 85%, an area under the curve (AUC) score of 93%, and an average precision (AP) score of 92%, promising diagnostic tool for cutaneous melanoma compared to traditional methods. Further improvements lay in the improvement of the architecture, expansion of the computational instances as well as of the dataset. Another field of future work could be devising a strategy for real-time implementation of this model in a hospital setting, as it could be of vital importance to provide swift support to doctors.O melanoma cutˆaneo ´e considerado o cancro de pele com a maior taxa de mortal idade e tem vindo a ganhar a aten¸c˜ao da comunidade m´edica devido ao seu r´apido aumento de incidˆencia. Os avan¸cos tecnol´ogicos contribu´ıram para m´etodos ino vadores de detec¸c˜ao de imagens transfer´ıveis para aplica¸c˜oes m´edicas, aumentando significativamente o potencial de interven¸c˜ao precoce no diagn´ostico de melanoma. Para tornar o diagn´ostico mais preciso e aumentar a taxa de sobrevivˆencia, este es tudo emprega t´ecnicas de aprendizagem profunda num conjunto alargado de dados provenientes de v´arias fontes. Utilizando a infraestrutura computacional Microsoft Azure Cloud, a abordagem de tentativa e erro foi utilizada ao hiperparametrizar v´arias redes neuronais convolucionais, sendo o crit´erio de decis˜ao a escolha daquela com a maior pontua¸c˜ao Fβ. MAR-MELA-CNN ´e um modelo ensemble que incor pora seis arquiteturas pr´e-treinadas: Xception, VGG16, ResNet50, NASNetMobile, MobileNetV2 e InceptionV3. O objetivo principal desta investiga¸c˜ao ´e potenciar a eficiˆencia das CNNs no diagn´ostico de melanoma e medir o seu desempenho num conjunto de dados unificado. O algoritmo proposto alcan¸cou uma pontua¸c˜ao Fβ de 85%, AUC de 93% e uma precis˜ao m´edia de 92%, tornando-se uma ferramenta promissora para o diagn´ostico de melanoma em compara¸c˜ao com os m´etodos tradi cionais. Os desenvolvimentos futuros incluem a melhoria da arquitetura e a extens˜ao das ferramentas computacionais e do conjunto de dados. Outro campo de trabalho futuro poderia ser a cria¸c˜ao de uma estrat´egia de implementa¸c˜ao em tempo real deste modelo num hospital, j´a que pode ser de vital importˆancia para fornecer apoio imediato aos m´edicos.Fernandes, Pedro AfonsoVeritati - Repositório Institucional da Universidade Católica PortuguesaSilva, Miguel Marinho Ferreira da2023-09-11T10:18:19Z2023-07-052023-062023-07-05T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10400.14/42247TID:203328868enginfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2023-09-19T01:41:54Zoai:repositorio.ucp.pt:10400.14/42247Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T20:28:58.859201Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Deep learning for melanoma classification : a study using skin lesion images |
title |
Deep learning for melanoma classification : a study using skin lesion images |
spellingShingle |
Deep learning for melanoma classification : a study using skin lesion images Silva, Miguel Marinho Ferreira da Cutaneous melanoma Deep learning Convolutional neural networks Fβ score Medical imaging Melanoma cutâneo Aprendizagem profunda Redes neuronais convolucionais Pontuação Fβ Imagiologia médica Domínio/Área Científica::Ciências Sociais::Economia e Gestão |
title_short |
Deep learning for melanoma classification : a study using skin lesion images |
title_full |
Deep learning for melanoma classification : a study using skin lesion images |
title_fullStr |
Deep learning for melanoma classification : a study using skin lesion images |
title_full_unstemmed |
Deep learning for melanoma classification : a study using skin lesion images |
title_sort |
Deep learning for melanoma classification : a study using skin lesion images |
author |
Silva, Miguel Marinho Ferreira da |
author_facet |
Silva, Miguel Marinho Ferreira da |
author_role |
author |
dc.contributor.none.fl_str_mv |
Fernandes, Pedro Afonso Veritati - Repositório Institucional da Universidade Católica Portuguesa |
dc.contributor.author.fl_str_mv |
Silva, Miguel Marinho Ferreira da |
dc.subject.por.fl_str_mv |
Cutaneous melanoma Deep learning Convolutional neural networks Fβ score Medical imaging Melanoma cutâneo Aprendizagem profunda Redes neuronais convolucionais Pontuação Fβ Imagiologia médica Domínio/Área Científica::Ciências Sociais::Economia e Gestão |
topic |
Cutaneous melanoma Deep learning Convolutional neural networks Fβ score Medical imaging Melanoma cutâneo Aprendizagem profunda Redes neuronais convolucionais Pontuação Fβ Imagiologia médica Domínio/Área Científica::Ciências Sociais::Economia e Gestão |
description |
Cutaneous melanoma is considered the skin cancer with highest mortality rate and has been gaining the attention of the medical community due to its rapidly increasing incidence. Advancements in computational technologies have paved the way for innovative image detection methods that can be transferable to medical ap plications, significantly enhancing the potential for early intervention in melanoma diagnosis. To make diagnosis more accurate and to further increase survival rates, this study employs deep learning techniques on an extensive dataset derived from multiple sources. Utilizing Microsoft Azure Cloud as the computational infras tructure, trial and error approach was employed by hyperparameterizing several convolutional neural networks (CNN) where the decision criteria were choosing the one with highest Fβ Score. MAR-MELA-CNN is an innovative ensemble model in corporating six state-of-the-art pre-trained CNN architectures: Xception, VGG16, ResNet50, NASNetMobile, MobileNetV2, and InceptionV3. The primary goal of this research is to further understand CNN’s efficiency in the diagnosis of melanoma and to furthermore measure its performance on a merged dataset. The proposed algorithm achieved a Fβ score of 85%, an area under the curve (AUC) score of 93%, and an average precision (AP) score of 92%, promising diagnostic tool for cutaneous melanoma compared to traditional methods. Further improvements lay in the improvement of the architecture, expansion of the computational instances as well as of the dataset. Another field of future work could be devising a strategy for real-time implementation of this model in a hospital setting, as it could be of vital importance to provide swift support to doctors. |
publishDate |
2023 |
dc.date.none.fl_str_mv |
2023-09-11T10:18:19Z 2023-07-05 2023-06 2023-07-05T00:00:00Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10400.14/42247 TID:203328868 |
url |
http://hdl.handle.net/10400.14/42247 |
identifier_str_mv |
TID:203328868 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
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.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799133557640134656 |