A methodology for developing dermatological datasets: lessons from retrospective data collection for AI-based applications
Our contribution offers a reproducible, scalable, and interdisciplinary framework for creating dermatological datasets, especially useful for countries initiating dataset creation. In addition to the methodological proposal, we highlight common pitfalls and offer recommendations to mitigate them.
Authors
- Leonel Hidalgo (ORCID: https://orcid.org/0000-0002-2782-2835)
- Ricardo Díaz-Domínguez (ORCID: https://orcid.org/0000-0001-5694-2679)
- Domingo Mery (ORCID: https://orcid.org/0000-0003-4748-3882)
- Ana M. Cabanas (ORCID: https://orcid.org/0000-0003-2800-150X)
- Cristián Navarrete‐Dechent (ORCID: https://orcid.org/0000-0003-4040-3640)
- José G. Tamez‐Peña (ORCID: https://orcid.org/0000-0003-1361-5162)
- Karina Carrasco (ORCID: https://orcid.org/0000-0001-5486-2464)
- Alma Pedro
- Pamela Romero
- Soledad Vidaurre
- Atsuko Galaz
Institutions
- Pontificia Universidad Católica de Chile (CL)
- University of Tarapacá (CL)
- Millennium Science Initiative (CL)
- Universidad Autónoma de Chile (CL)
- Millennium Institute for Integrative Biology (CL)
- Tecnológico de Monterrey (MX)
Publication Details
- Journal
- BMC Medical Research Methodology
- Published
- 2025-11-05
- DOI
- https://doi.org/10.1186/s12874-025-02706-y
- Primary Topic
- Cutaneous Melanoma Detection and Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Institut Català de Nanociència i Nanotecnologia
- Agenția Națională pentru Cercetare și Dezvoltare
- Agencia Nacional de Investigación y Desarrollo
- CHIST-ERA