Cargando…

Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral

Teledermatology has developed rapidly in recent years and is nowadays an essential tool for early diagnosis. In this work, we aim to improve existing Teledermatology processes for skin lesion diagnosis by developing a deep learning approach for risk prioritization with a dataset of retrospective dat...

Descripción completa

Detalles Bibliográficos
Autores principales: Carvalho, Rafaela, Morgado, Ana C., Andrade, Catarina, Nedelcu, Tudor, Carreiro, André, Vasconcelos, Maria João M.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8775114/
https://www.ncbi.nlm.nih.gov/pubmed/35054203
http://dx.doi.org/10.3390/diagnostics12010036
_version_ 1784636505182961664
author Carvalho, Rafaela
Morgado, Ana C.
Andrade, Catarina
Nedelcu, Tudor
Carreiro, André
Vasconcelos, Maria João M.
author_facet Carvalho, Rafaela
Morgado, Ana C.
Andrade, Catarina
Nedelcu, Tudor
Carreiro, André
Vasconcelos, Maria João M.
author_sort Carvalho, Rafaela
collection PubMed
description Teledermatology has developed rapidly in recent years and is nowadays an essential tool for early diagnosis. In this work, we aim to improve existing Teledermatology processes for skin lesion diagnosis by developing a deep learning approach for risk prioritization with a dataset of retrospective data from referral requests of the Portuguese National Health System. Given the high complexity of this task, we propose a new prioritization pipeline guided and inspired by domain knowledge. We explored automatic lesion segmentation and tested different learning schemes, namely hierarchical classification and curriculum learning approaches, optionally including additional patient metadata. The final priority level prediction can then be obtained by combining predicted diagnosis and a baseline priority level accounting for explicit expert knowledge. In both the differential diagnosis and prioritization branches, lesion segmentation with 30% tolerance for contextual information was shown to improve classification when compared with a flat baseline model trained on original images; furthermore, the addition of patient information was not beneficial for most experiments. Curriculum learning delivered better results than a flat or hierarchical approach. The combination of diagnosis information and a knowledge map, created in collaboration with dermatologists, together with the priority achieved interesting results (best macro F1 of 43.93% for a validated test set), paving the way for new data-centric and knowledge-driven approaches.
format Online
Article
Text
id pubmed-8775114
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-87751142022-01-21 Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral Carvalho, Rafaela Morgado, Ana C. Andrade, Catarina Nedelcu, Tudor Carreiro, André Vasconcelos, Maria João M. Diagnostics (Basel) Article Teledermatology has developed rapidly in recent years and is nowadays an essential tool for early diagnosis. In this work, we aim to improve existing Teledermatology processes for skin lesion diagnosis by developing a deep learning approach for risk prioritization with a dataset of retrospective data from referral requests of the Portuguese National Health System. Given the high complexity of this task, we propose a new prioritization pipeline guided and inspired by domain knowledge. We explored automatic lesion segmentation and tested different learning schemes, namely hierarchical classification and curriculum learning approaches, optionally including additional patient metadata. The final priority level prediction can then be obtained by combining predicted diagnosis and a baseline priority level accounting for explicit expert knowledge. In both the differential diagnosis and prioritization branches, lesion segmentation with 30% tolerance for contextual information was shown to improve classification when compared with a flat baseline model trained on original images; furthermore, the addition of patient information was not beneficial for most experiments. Curriculum learning delivered better results than a flat or hierarchical approach. The combination of diagnosis information and a knowledge map, created in collaboration with dermatologists, together with the priority achieved interesting results (best macro F1 of 43.93% for a validated test set), paving the way for new data-centric and knowledge-driven approaches. MDPI 2021-12-24 /pmc/articles/PMC8775114/ /pubmed/35054203 http://dx.doi.org/10.3390/diagnostics12010036 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Carvalho, Rafaela
Morgado, Ana C.
Andrade, Catarina
Nedelcu, Tudor
Carreiro, André
Vasconcelos, Maria João M.
Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title_full Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title_fullStr Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title_full_unstemmed Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title_short Integrating Domain Knowledge into Deep Learning for Skin Lesion Risk Prioritization to Assist Teledermatology Referral
title_sort integrating domain knowledge into deep learning for skin lesion risk prioritization to assist teledermatology referral
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8775114/
https://www.ncbi.nlm.nih.gov/pubmed/35054203
http://dx.doi.org/10.3390/diagnostics12010036
work_keys_str_mv AT carvalhorafaela integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral
AT morgadoanac integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral
AT andradecatarina integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral
AT nedelcutudor integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral
AT carreiroandre integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral
AT vasconcelosmariajoaom integratingdomainknowledgeintodeeplearningforskinlesionriskprioritizationtoassistteledermatologyreferral