Browsing by Author "Cabanas, Ana M."

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    A methodology for developing dermatological datasets: lessons from retrospective data collection for AI-based applications
    (2025) Pedro Pérez, Alma Alheli; Romero Jofré, Pamela Ignacia; Vidaurre, Soledad; Cabanas, Ana M.; Galaz, Atsuko; Hidalgo Acuña, Leonel Esteban; Carrasco, Karina; Tamez-Peña, José Gerardo; Díaz-Domínguez, Ricardo; Navarrete Dechent, Cristian Patricio; Mery Quiroz, Domingo Arturo
    Purpose The integration of artificial intelligence into dermatological research has underscored the need for robust and well-structured dermatological datasets. However, these datasets vary widely in their development processes, and there is currently no standard methodology to create such datasets. This work identifies three pressing needs for the building of dermatological datasets focus on skin tumor classification: the need for multimodal datasets, the definition of minimum metadata requirements, and the inclusion of underrepresented populations to address the scarcity of health data. Methods We propose a practical methodology to create dermatological datasets from clinical records, incorporating both images and patient metadata. The process consists of four key stages: getting the institutional review board approval and analysis of clinical information sources, data recording and structuring, processing of clinical data and images, and quality assessment. This methodology was derived from hands-on experience in building two datasets from Chilean and Mexican populations, respectively. Results The methodology allows the creation of well-structured datasets by simplifying data organization and enabling replication. Each step includes practical guidance for dealing with typical challenges, such as image metadata categorization and technical validation by dermatologists and computer scientists. Conclusion 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.
  • No Thumbnail Available
    Item
    Technical and regulatory challenges in artificial intelligence-based pulse oximetry: a proposed development pipeline
    (Elsevier Ltd., 2025) Cabanas, Ana M.; Martín-Escudero, Pilar; Pagan, Josué; Mery Quiroz, Domingo
    Pulse oximetry, although generally effective under ideal conditions, faces challenges in accurately estimating peripheral oxygen saturation (SpO2) in complex clinical scenarios, particularly at lower saturation levels and in patients with darker skin pigmentation. Artificial intelligence (AI) offers the potential to improve SpO2 monitoring by enabling more accurate, equitable, and accessible estimations. We highlight key challenges in developing AI-enhanced pulse oximetry, including the need for diverse and representative datasets, refined validation protocols addressing ethical concerns such as algorithmic bias, expanded SpO2 measurement ranges encompassing hypoxaemic levels, and enhanced model interpretability. We emphasise the importance of transitioning from subjective skin tone assessments to quantitative methods to ensure equity and mitigate bias. Finally, we propose a development pipeline and discuss strategies for robust, fair AI-based SpO2 monitoring, including aligning validation with global regulatory frameworks and fostering interdisciplinary collaboration. These advances will improve the reliability and fairness of pulse oximetry, ultimately contributing to enhanced global patient care.