Prediction of melanoma Breslow thickness using deep transfer learning algorithms
Clinical and Experimental Dermatology · 07.07.2023 · 48 (7) · 752-758
DOI: 10.1093/ced/llad107 ↗Kurz gefasst
Fachartikel in der Zeitschrift Clinical and Experimental Dermatology, online erschienen am 27.03.2023. Verfasst von Juan-Carlos Hernández-Rodríguez und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Dermatology, Virgen del Rocio University Hospital, Seville, Spain. Hauptthemen laut MEDLINE: Skin Neoplasms, Melanoma. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 26 Literaturangaben.
Abstract (Originalsprache)
Background
The distinction between in situ melanoma (MIS) and invasive melanoma is challenging even for expert dermatologists. The use of pretrained convolutional neural networks (CNNs) as ancillary decision systems needs further research.
Aim
To develop, validate and compare three deep transfer learning (DTL) algorithms to predict MIS vs. invasive melanoma and melanoma with a Breslow thickness (BT) of < 0.8 mm vs. ≥ 0.8 mm.
Methods
A dataset of 1315 dermoscopic images of histopathologically confirmed melanomas was created from Virgen del Rocio University Hospital and open repositories of the International Skin Imaging Collaboration archive and Polesie S et al. (Dermatol Pract Concept 2021; 11:e2021079). The images were labelled as MIS or invasive melanoma and < 0.8 mm or ≥ 0.8 mm of BT. We conducted three trainings, and overall means for receiver operating characteristic (ROC) curves, sensitivity, specificity, positive and negative predictive value, and balanced diagnostic accuracy outcomes were evaluated on the test set with ResNetV2, EfficientNetB6 and InceptionV3. The results of 10 dermatologists were compared with the algorithms. Grad-CAM gradient maps were generated, highlighting relevant areas considered by the CNNs within the images.
Results
EfficientNetB6 achieved the highest diagnostic accuracy for the comparison between MIS vs. invasive melanoma (61%) and BT < 0.8 mm vs. ≥ 0.8 mm (75%). For the BT comparison, ResNetV2 with an area under the ROC curve of 0.76 and InceptionV3 with an area under the ROC curve of 0.75, outperformed the results obtained by the dermatologist group with an area under the ROC curve of 0.70.
Conclusion
EfficientNetB6 recorded the best prediction results, outperforming the dermatologists for the comparison of 0.8 mm of BT. DTL could be an ancillary aid to support dermatologists’ decisions in the near future.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Dermatology, Virgen del Rocio University Hospital, Seville, Spain.
- Department of Medicine, Faculty of Medicine.
- Robotics and Technology of Computers Laboratory, University of Seville, Seville, Spain.
ORCID-Profile: Juan-Carlos Hernández-Rodríguez, Lourdes Durán-López, Juan P Domínguez-Morales, Juan Ortiz-Álvarez, Julián Conejo-Mir, Jose-Juan Pereyra-Rodriguez
Zitieren
APA
Hernández-Rodríguez, J. C., Durán-López, L., Domínguez-Morales, J. P., Ortiz-Álvarez, J., Conejo-Mir, J., & Pereyra-Rodriguez, J. J. (2023). Prediction of melanoma Breslow thickness using deep transfer learning algorithms. Clinical and Experimental Dermatology, 48(7), 752-758. https://doi.org/10.1093/ced/llad107
Vancouver
Hernández-Rodríguez JC, Durán-López L, Domínguez-Morales JP, Ortiz-Álvarez J, Conejo-Mir J, Pereyra-Rodriguez JJ. Prediction of melanoma Breslow thickness using deep transfer learning algorithms. Clinical and Experimental Dermatology. 2023;48(7):752-758. doi:10.1093/ced/llad107
BibTeX
@article{HernndezRodrguez2023,
author = {Hernández-Rodríguez, Juan-Carlos and Durán-López, Lourdes and Domínguez-Morales, Juan P and Ortiz-Álvarez, Juan and Conejo-Mir, Julián and Pereyra-Rodriguez, Jose-Juan},
title = {Prediction of melanoma Breslow thickness using deep transfer learning algorithms},
journal = {Clinical and Experimental Dermatology},
year = {2023},
volume = {48},
number = {7},
pages = {752--758},
doi = {10.1093/ced/llad107}
}
EUVIMED