EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

EffiSegNet : Advances in Automated Segmentation of Actinic Cheilitis and Lip Squamous Cell Carcinoma in Clinical Photographs

Vitor Simbalista Teixeira Soares iD, Anna Luíza Damaceno Araújo iD, Ivan José Correia-Neto iD, Cristina Saldivia-Siracusa iD, Raísa Sales de Sá iD, Hugo Abritta Reis iD, Ayrton da Costa Ganem Filho iD, André Carlos Ponce de Leon Ferreira de Carvalho iD, Thierry Urruty iD, Paulo Rogério Ferreti Bonan iD, Nathalia Rodrigues Gomes iD, Felipe Paiva Fonseca iD und 5 weitere

Journal of Oral Pathology & Medicine · 2026-10 · 55 (9) · 1073-1080

DOI: 10.1111/jop.70176 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Oral Pathology & Medicine, online erschienen am 23.08.2026. Verfasst von Vitor Simbalista Teixeira Soares und 16 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Science and Technology (ICT-UNIFESP), Federal University of São Paulo, São José dos Campos, São Paulo, Brazil. Hauptthemen laut MEDLINE: Cheilitis, Lip Neoplasms, Carcinoma, Squamous Cell, Deep Learning. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 14 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: São Paulo Research Foundation (FAPESP), Brazil, National Council for Scientific and Technological Development, Minas Gerais Research Foundation (FAPEMIG).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Actinic cheilitis (AC) is a potentially malignant oral disorder linked to lip squamous cell carcinoma (LSCC). Clinical diagnosis is hindered by lesion heterogeneity. While AI tools have shown promise in oral lesion detection and diagnosis, image segmentation remains underexplored. This study developed and externally validated a deep learning segmentation model for digital photographs of lip lesions.

Methods

A total of 489 clinical photographs from one institution were used for training/internal testing, and 82 external images from two independent institutions (UFPB and UFMG) served as external validation. EffiSegNet‑B0 used Image Net‑pretrained EfficientNet‑B0 encoder, with stage features aggregated via Ghost Modules. Manual segmentation masks comprised the clinically affected lip region and were used as ground truth. Performance was assessed using accuracy, sensitivity (recall), precision, Dice Similarity Coefficient (DSC) and intersection over union (IoU).

Results

The model achieved high accuracy (> 95%) on both internal and external datasets. DSCs were 0.888 (95% CI: 0.871–0.904) for internal test and 0.875 (95% CI: 0.862–0.888) for external validation, while IoU values were 0.802 (95% CI: 0.777–0.828) and 0.782 (95% CI: 0.763–0.801), respectively. No performance decline was observed on external validation data. The model remained robust to lesion variations (ulceration, crusting, and keratinization).

Conclusion

EffiSegNet‐B0 demonstrated high accuracy and adaptability in segmenting lip lesions, reinforcing its potential for clinical image analysis applications. The proposed model generalizes well across different institutions and lesion phenotypes, consistently extracting the complete region of interest. This proof‐of‐concept study supports its promise as a tool for automated lip lesion segmentation in heterogeneous clinical settings.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

CheilitisLip NeoplasmsCarcinoma, Squamous CellDeep LearningImage Processing, Computer-AssistedPhotographyHumansLip

Stichwörter der Autor:innen

actinic cheilitisartificial intelligenceconvolutional neural networkdeep learninglip squamous cell carcinomasegmentation

Autor:innen und Institutionen

  • Institute of Science and Technology (ICT-UNIFESP), Federal University of São Paulo, São José dos Campos, São Paulo, Brazil.
  • Head and Neck Surgery Department, University of São Paulo Medical School (FMUSP), São Paulo, Brazil.
  • Hospital Israelita Albert Einstein, São Paulo, Brazil.
  • Departamento de Diagnóstico Oral, Faculdade de Odontologia de Piracicaba, Universidade Estadual de Campinas (FOP-UNICAMP), Piracicaba, São Paulo, Brazil.
  • School of Dentistry, Ceuma University, São Luís, Brazil.
  • Institute of Mathematics and Computer Sciences of University of São Paulo (ICMC-USP), São Paulo, Brazil.

ORCID-Profile: Vitor Simbalista Teixeira Soares, Anna Luíza Damaceno Araújo, Ivan José Correia-Neto, Cristina Saldivia-Siracusa, Raísa Sales de Sá, Hugo Abritta Reis, Ayrton da Costa Ganem Filho, André Carlos Ponce de Leon Ferreira de Carvalho, Thierry Urruty, Paulo Rogério Ferreti Bonan, Nathalia Rodrigues Gomes, Felipe Paiva Fonseca, Pablo Agustin Vargas, Alan Roger Santos-Silva, Luiz Paulo Kowalski, Matheus Cardoso Moraes, Marcio Ajudarte Lopes

Förderung

  • São Paulo Research Foundation (FAPESP), Brazil
  • National Council for Scientific and Technological Development
  • Minas Gerais Research Foundation (FAPEMIG)
  • Fundação de Amparo à Pesquisa do Estado de São Paulo
  • Conselho Nacional de Desenvolvimento Científico e Tecnológico
  • Fundação de Amparo à Pesquisa do Estado de Minas Gerais

Zitieren

APA

Soares, V. S. T., Araújo, A. L. D., Correia-Neto, I. J., Saldivia-Siracusa, C., Sá, R. S., Reis, H. A., Filho, A. D. C. G., Carvalho, A. C. P. L. F., Urruty, T., Bonan, P. R. F., Gomes, N. R., Fonseca, F. P., Vargas, P. A., Santos-Silva, A. R., Kowalski, L. P., Moraes, M. C., & Lopes, M. A. (2026). EffiSegNet : Advances in Automated Segmentation of Actinic Cheilitis and Lip Squamous Cell Carcinoma in Clinical Photographs. Journal of Oral Pathology & Medicine, 55(9), 1073-1080. https://doi.org/10.1111/jop.70176

Vancouver

Soares VST, Araújo ALD, Correia-Neto IJ, Saldivia-Siracusa C, Sá RS, Reis HA, et al. EffiSegNet : Advances in Automated Segmentation of Actinic Cheilitis and Lip Squamous Cell Carcinoma in Clinical Photographs. Journal of Oral Pathology & Medicine. 2026;55(9):1073-1080. doi:10.1111/jop.70176

BibTeX
@article{Soares2026,
  author = {Soares, Vitor Simbalista Teixeira and Araújo, Anna Luíza Damaceno and Correia-Neto, Ivan José and Saldivia-Siracusa, Cristina and Sá, Raísa Sales de and Reis, Hugo Abritta and Filho, Ayrton da Costa Ganem and Carvalho, André Carlos Ponce de Leon Ferreira de and Urruty, Thierry and Bonan, Paulo Rogério Ferreti and Gomes, Nathalia Rodrigues and Fonseca, Felipe Paiva and Vargas, Pablo Agustin and Santos-Silva, Alan Roger and Kowalski, Luiz Paulo and Moraes, Matheus Cardoso and Lopes, Marcio Ajudarte},
  title = {EffiSegNet : Advances in Automated Segmentation of Actinic Cheilitis and Lip Squamous Cell Carcinoma in Clinical Photographs},
  journal = {Journal of Oral Pathology & Medicine},
  year = {2026},
  volume = {55},
  number = {9},
  pages = {1073--1080},
  doi = {10.1111/jop.70176}
}
Kostenlos starten →