Artificial intelligence algorithm for the histopathological diagnosis of skin cancer
Autor(a) principal: | |
---|---|
Data de Publicação: | 2021 |
Outros Autores: | , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Clinical and Biomedical Research |
Texto Completo: | https://seer.ufrgs.br/index.php/hcpa/article/view/108236 |
Resumo: | Introduction: Cutaneous neoplasms are the most common cancers in the world, and have high morbidity rates. A definitive diagnosis can only be obtained after histopathological evaluation of the lesions. Objective: To develop an artificial intelligence program to establish the histopathological diagnosis of cutaneous lesions. Methodology: A deep learning program was built using three neural network architectures: MobileNet, Inception and convolutional networks. A database was constructed using 2732 images of melanomas, basal and squamous cell carcinomas, and normal skin. The validation set consisted of 284 images from all 4 categories, allowing for the calculation of sensitivity and specificity. All images were provided by the Path Presenter website. Results: The sensitivity and specificity of the MobileNet model were 92% (95%CI, 83-100%) and 97% (95%CI, 90-100%), respectively; corresponding figures for the Inception model were 98.3% (95%CI, 86-100%) and 98.8% (95%CI, 98.2-100%); lastly, the sensitivity and specificity of the convolutional network model were 91.6% (95%CI, 73.8-100%) and 95.7% (95%CI, 94.4-97.2%). The maximum sensitivity for the differentiation of malignant conditions was 91%, and specificity was 95.4%. Conclusion: The program developed in the present study can efficiently distinguish between the main types of skin cancer with high sensitivity and specificity. |
id |
UFRGS-20_bd40e209b522a849ae0bb7ad8f315088 |
---|---|
oai_identifier_str |
oai:seer.ufrgs.br:article/108236 |
network_acronym_str |
UFRGS-20 |
network_name_str |
Clinical and Biomedical Research |
repository_id_str |
|
spelling |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancerSkin lesionsArtificial IntelligenceDiagnosisNeoplasmsMelanoma.Skin woundArtificial IntelligenceDiagnosisNeoplasmsMelanomaIntroduction: Cutaneous neoplasms are the most common cancers in the world, and have high morbidity rates. A definitive diagnosis can only be obtained after histopathological evaluation of the lesions. Objective: To develop an artificial intelligence program to establish the histopathological diagnosis of cutaneous lesions. Methodology: A deep learning program was built using three neural network architectures: MobileNet, Inception and convolutional networks. A database was constructed using 2732 images of melanomas, basal and squamous cell carcinomas, and normal skin. The validation set consisted of 284 images from all 4 categories, allowing for the calculation of sensitivity and specificity. All images were provided by the Path Presenter website. Results: The sensitivity and specificity of the MobileNet model were 92% (95%CI, 83-100%) and 97% (95%CI, 90-100%), respectively; corresponding figures for the Inception model were 98.3% (95%CI, 86-100%) and 98.8% (95%CI, 98.2-100%); lastly, the sensitivity and specificity of the convolutional network model were 91.6% (95%CI, 73.8-100%) and 95.7% (95%CI, 94.4-97.2%). The maximum sensitivity for the differentiation of malignant conditions was 91%, and specificity was 95.4%. Conclusion: The program developed in the present study can efficiently distinguish between the main types of skin cancer with high sensitivity and specificity.HCPA/FAMED/UFRGS2021-04-13info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionPeer-reviewed ArticleAvaliado por Paresapplication/pdfhttps://seer.ufrgs.br/index.php/hcpa/article/view/108236Clinical & Biomedical Research; Vol. 40 No. 4 (2020): Clinical and Biomedical ResearchClinical and Biomedical Research; v. 40 n. 4 (2020): Clinical and Biomedical Research2357-9730reponame:Clinical and Biomedical Researchinstname:Universidade Federal do Rio Grande do Sul (UFRGS)instacron:UFRGSenghttps://seer.ufrgs.br/index.php/hcpa/article/view/108236/pdfCopyright (c) 2021 Clinical and Biomedical Researchinfo:eu-repo/semantics/openAccessKuiava, Victor AntonioKuiava, Eliseu LuizChielle, Eduardo OttobelliDe Bittencourt, Francisco Madalosso2024-01-19T14:20:54Zoai:seer.ufrgs.br:article/108236Revistahttps://www.seer.ufrgs.br/index.php/hcpaPUBhttps://seer.ufrgs.br/index.php/hcpa/oai||cbr@hcpa.edu.br2357-97302357-9730opendoar:2024-01-19T14:20:54Clinical and Biomedical Research - Universidade Federal do Rio Grande do Sul (UFRGS)false |
dc.title.none.fl_str_mv |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
title |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
spellingShingle |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer Kuiava, Victor Antonio Skin lesions Artificial Intelligence Diagnosis Neoplasms Melanoma. Skin wound Artificial Intelligence Diagnosis Neoplasms Melanoma |
title_short |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
title_full |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
title_fullStr |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
title_full_unstemmed |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
title_sort |
Artificial intelligence algorithm for the histopathological diagnosis of skin cancer |
author |
Kuiava, Victor Antonio |
author_facet |
Kuiava, Victor Antonio Kuiava, Eliseu Luiz Chielle, Eduardo Ottobelli De Bittencourt, Francisco Madalosso |
author_role |
author |
author2 |
Kuiava, Eliseu Luiz Chielle, Eduardo Ottobelli De Bittencourt, Francisco Madalosso |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Kuiava, Victor Antonio Kuiava, Eliseu Luiz Chielle, Eduardo Ottobelli De Bittencourt, Francisco Madalosso |
dc.subject.por.fl_str_mv |
Skin lesions Artificial Intelligence Diagnosis Neoplasms Melanoma. Skin wound Artificial Intelligence Diagnosis Neoplasms Melanoma |
topic |
Skin lesions Artificial Intelligence Diagnosis Neoplasms Melanoma. Skin wound Artificial Intelligence Diagnosis Neoplasms Melanoma |
description |
Introduction: Cutaneous neoplasms are the most common cancers in the world, and have high morbidity rates. A definitive diagnosis can only be obtained after histopathological evaluation of the lesions. Objective: To develop an artificial intelligence program to establish the histopathological diagnosis of cutaneous lesions. Methodology: A deep learning program was built using three neural network architectures: MobileNet, Inception and convolutional networks. A database was constructed using 2732 images of melanomas, basal and squamous cell carcinomas, and normal skin. The validation set consisted of 284 images from all 4 categories, allowing for the calculation of sensitivity and specificity. All images were provided by the Path Presenter website. Results: The sensitivity and specificity of the MobileNet model were 92% (95%CI, 83-100%) and 97% (95%CI, 90-100%), respectively; corresponding figures for the Inception model were 98.3% (95%CI, 86-100%) and 98.8% (95%CI, 98.2-100%); lastly, the sensitivity and specificity of the convolutional network model were 91.6% (95%CI, 73.8-100%) and 95.7% (95%CI, 94.4-97.2%). The maximum sensitivity for the differentiation of malignant conditions was 91%, and specificity was 95.4%. Conclusion: The program developed in the present study can efficiently distinguish between the main types of skin cancer with high sensitivity and specificity. |
publishDate |
2021 |
dc.date.none.fl_str_mv |
2021-04-13 |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Peer-reviewed Article Avaliado por Pares |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
https://seer.ufrgs.br/index.php/hcpa/article/view/108236 |
url |
https://seer.ufrgs.br/index.php/hcpa/article/view/108236 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
https://seer.ufrgs.br/index.php/hcpa/article/view/108236/pdf |
dc.rights.driver.fl_str_mv |
Copyright (c) 2021 Clinical and Biomedical Research info:eu-repo/semantics/openAccess |
rights_invalid_str_mv |
Copyright (c) 2021 Clinical and Biomedical Research |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
HCPA/FAMED/UFRGS |
publisher.none.fl_str_mv |
HCPA/FAMED/UFRGS |
dc.source.none.fl_str_mv |
Clinical & Biomedical Research; Vol. 40 No. 4 (2020): Clinical and Biomedical Research Clinical and Biomedical Research; v. 40 n. 4 (2020): Clinical and Biomedical Research 2357-9730 reponame:Clinical and Biomedical Research instname:Universidade Federal do Rio Grande do Sul (UFRGS) instacron:UFRGS |
instname_str |
Universidade Federal do Rio Grande do Sul (UFRGS) |
instacron_str |
UFRGS |
institution |
UFRGS |
reponame_str |
Clinical and Biomedical Research |
collection |
Clinical and Biomedical Research |
repository.name.fl_str_mv |
Clinical and Biomedical Research - Universidade Federal do Rio Grande do Sul (UFRGS) |
repository.mail.fl_str_mv |
||cbr@hcpa.edu.br |
_version_ |
1799767056207314944 |