EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Feasibility study of an AI-powered mobile app to support cutaneous leishmaniasis diagnosis in the Brazilian Amazon

Karina Lumy Okita iD, Thamires Bastos Pinheiro, Livia Oliveira-Ciabati, Breno Dos Santos Alves, Pedro Luis Gil Bonett, Elinea Liborio Fragoso, Eddy Jens Rivero-Zavala, Antonio Joaquim Fernandes, Gustavo Costa Souza, Jorge Ewerton Dos Santos Sales, Maria das Graças Vale Barbosa Guerra, Jorge Augusto de Oliveira Guerra und 1 weitere

PLOS Neglected Tropical Diseases · 27.05.2026 · 20 (5) · e0014313

DOI: 10.1371/journal.pntd.0014313 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS Neglected Tropical Diseases, online erschienen am 27.05.2026. Verfasst von Karina Lumy Okita und 12 weiteren Autor:innen. Beteiligte Einrichtung: Health Innovation Techcenter, Hospital Israelita Albert Einstein: Sociedade Beneficente Israelita Brasileira Albert Einstein, São Paulo, São Paulo, B…. Hauptthemen laut MEDLINE: Leishmaniasis, Cutaneous, Mobile Applications, Artificial Intelligence. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Inter-American Development Bank.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Cutaneous leishmaniasis (CL) remains a major public health challenge, especially in Brazil’s Amazon, where environmental and economic pressures sustain transmission. Delayed diagnosis drives morbidity, stigma, and costs. Community health workers are pivotal yet under-equipped for early triage. Artificial intelligence, effective in dermatologic imaging, is underused for CL; feasible, offline clinical tools could accelerate referral and timely care. This study aimed to evaluate the feasibility of an AI-assisted tool for early CL triage in the Brazilian Amazon by: (i) developing an AI model for CL identification from clinical skin lesion images; (ii) integrating the model into an offline-optimized mobile application for resource-limited settings; and (iii) conducting initial, real-world clinical validation.

Methods

Exploratory, IRB-approved feasibility study. Retrospective images from Brazilian Amazon sites informed a two-stage AI pipeline (lesion segmentation+classification) integrated into an offline mobile app. Multicenter validation in ideal- and real-world scenarios. Primary metric AUC-ROC; secondary sensitivity/specificity. One-sided tests: AUC-ROC > 0.70; ideal-world sensitivity > 0.75.

Results

For the Classification Model, 64 images were assigned to the test set while 1,160 were used for training and validation (80:20 split), with DenseNet121 yielding the highest accuracy of 0.88. The full AI pipeline (Segmentation and Classification models) achieved an accuracy of 0.81, an F1-score of 0.80, an AUC-ROC of 0.90, and a sensitivity of 0.76. In the ideal-world analysis, sensitivity reached 0.92, the F1-score was 0.84, and specificity was 0.42.

Conclusions

We demonstrate the feasibility of an offline, AI-assisted mobile tool to support triage and referral for cutaneous leishmaniasis in the Brazilian Amazon. Performance reflects preliminary, descriptive point estimates from an initial diagnostic accuracy assessment and should be interpreted with caution; the tool is not intended for standalone diagnosis. Next steps include prospectively powered clinical validation, usability refinements, and regulatory evaluation, including alignment with ANVISA requirements in Brazil. Overall, this work represents an initial step toward closing the gap in clinically supported diagnostic tools for neglected tropical diseases in resource-constrained settings.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Leishmaniasis, CutaneousMobile ApplicationsArtificial IntelligenceHumansFeasibility StudiesBrazilFemaleSensitivity and SpecificityMaleAdultRetrospective StudiesAdolescentYoung AdultMiddle AgedChildSkin

Autor:innen und Institutionen

  • Health Innovation Techcenter, Hospital Israelita Albert Einstein: Sociedade Beneficente Israelita Brasileira Albert Einstein, São Paulo, São Paulo, Brasil.
  • Dermatology Department, Universidade de Mogi das Cruzes, Mogi das Cruzes, São Paulo, Brasil.
  • Program in Applied Sciences in Dermatology at the State University of Amazonas, Manaus, Amazonas, Brasil.
  • Centro Universitário Barão de Mauá, Ribeirão Preto, São Paulo, Brasil.
  • Fundação Hospitalar Alfredo da Matta, Manaus, Amazonas, Brasil.
  • Fundação de Medicina Tropical Doutor Heitor Vieira Dourado, Manaus, Amazonas, Brasil.

ORCID-Profile: Karina Lumy Okita

Förderung

  • Inter-American Development Bank

Zitieren

APA

Okita, K. L., Pinheiro, T. B., Oliveira-Ciabati, L., Alves, B. D. S., Bonett, P. L. G., Fragoso, E. L., Rivero-Zavala, E. J., Fernandes, A. J., Souza, G. C., Sales, J. E. D. S., Guerra, M. D. G. V. B., Guerra, J. A. O., & Carvalho, I. (2026). Feasibility study of an AI-powered mobile app to support cutaneous leishmaniasis diagnosis in the Brazilian Amazon. PLOS Neglected Tropical Diseases, 20(5), e0014313. https://doi.org/10.1371/journal.pntd.0014313

Vancouver

Okita KL, Pinheiro TB, Oliveira-Ciabati L, Alves BDS, Bonett PLG, Fragoso EL, et al. Feasibility study of an AI-powered mobile app to support cutaneous leishmaniasis diagnosis in the Brazilian Amazon. PLOS Neglected Tropical Diseases. 2026;20(5):e0014313. doi:10.1371/journal.pntd.0014313

BibTeX
@article{Okita2026,
  author = {Okita, Karina Lumy and Pinheiro, Thamires Bastos and Oliveira-Ciabati, Livia and Alves, Breno Dos Santos and Bonett, Pedro Luis Gil and Fragoso, Elinea Liborio and Rivero-Zavala, Eddy Jens and Fernandes, Antonio Joaquim and Souza, Gustavo Costa and Sales, Jorge Ewerton Dos Santos and Guerra, Maria das Graças Vale Barbosa and Guerra, Jorge Augusto de Oliveira and Carvalho, Isabelle},
  title = {Feasibility study of an AI-powered mobile app to support cutaneous leishmaniasis diagnosis in the Brazilian Amazon},
  journal = {PLOS Neglected Tropical Diseases},
  year = {2026},
  volume = {20},
  number = {5},
  pages = {e0014313},
  doi = {10.1371/journal.pntd.0014313}
}
Kostenlos starten →