A Blockchain-Integrated Pipeline Featuring Generative AI for Melanoma Detection, Segmentation, and Classification in Dermoscopic Images
Journal of the Brazilian Computer Society · 03.07.2026 · 32 (1) · 1786-1811
DOI: 10.5753/jbcs.2026.5903 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the Brazilian Computer Society, online erschienen am 03.07.2026. Verfasst von Marcus Vinicius Candido de Figueiredo und 5 weiteren Autor:innen. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Melanoma is one of the most aggressive forms of skin cancer, responsible for thousands of deaths annually worldwide. The effectiveness of early diagnosis and continuous monitoring of disease progression are critical factors for therapeutic success and the improvement of public health outcomes. This study presents an advanced blockchain-integrated pipeline for the automated analysis of dermoscopic images, combining modern Computer Vision and Artificial Intelligence (AI) techniques to enhance the processes of lesion detection, segmentation, and classification. Using state-of-the-art architectures such as YOLO for identifying suspicious regions and the Segment Anything Model (SAM) for precise extraction of affected areas, the proposed method achieved an accuracy of 99% in the initial analysis stages, outperforming previously reported results in the literature. For the classification stage, a pipeline of feature extraction and selection is employed, enabling the identification of regions with a high probability of malignancy, achieving 86% accuracy. The study also incorporates Large Language Models (LLMs) to automatically generate interpretive reports, providing preliminary clinical decision support through generative AI. In the final stage, all generated data and results—including processed images, model outputs, and generated reports—are recorded on a blockchain network. This distributed storage ensures the integrity, traceability, and immutability of the information, enabling the creation of a secure and auditable clinical history. The blockchain-based solution adds transparency to the proposed method and enables controlled data sharing among medical institutions, reinforcing trust and accountability in the use of automated tools for dermatological diagnosis support.
Zitieren
APA
Figueiredo, M. V. C. D., Marques, A. G., Silva, F. I. G. D., Nascimento, J. J. D. C., Souza, C. T. D., & Souza, L. F. D. F. (2026). A Blockchain-Integrated Pipeline Featuring Generative AI for Melanoma Detection, Segmentation, and Classification in Dermoscopic Images. Journal of the Brazilian Computer Society, 32(1), 1786-1811. https://doi.org/10.5753/jbcs.2026.5903
Vancouver
Figueiredo MVCD, Marques AG, Silva FIGD, Nascimento JJDC, Souza CTD, Souza LFDF. A Blockchain-Integrated Pipeline Featuring Generative AI for Melanoma Detection, Segmentation, and Classification in Dermoscopic Images. Journal of the Brazilian Computer Society. 2026;32(1):1786-1811. doi:10.5753/jbcs.2026.5903
BibTeX
@article{Figueiredo2026,
author = {Figueiredo, Marcus Vinicius Candido de and Marques, Adriell Gomes and Silva, Francisco Italo Guilherme Da and Nascimento, José Jerovane Da Costa and Souza, Cidcley Teixeira de and Souza, Luís Fabrício De Freitas},
title = {A Blockchain-Integrated Pipeline Featuring Generative AI for Melanoma Detection, Segmentation, and Classification in Dermoscopic Images},
journal = {Journal of the Brazilian Computer Society},
year = {2026},
volume = {32},
number = {1},
pages = {1786--1811},
doi = {10.5753/jbcs.2026.5903}
}
EUVIMED