Diagnóstico de melanoma cutáneo usando redes neuronales convolucionales en teléfonos móviles
This research develops and validates a mobile application based on convolutional neural networks (CNN) for the diagnosis of cutaneous melanoma, in order to offer an accessible and accurate tool that facilitates early detection in non-specialized users. The problem addressed is the limited accessibil...
| Autores: | , |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2025 |
| País: | Perú |
| Institución: | Universidad Nacional De La Amazonía Peruana |
| Repositorio: | UNAPIquitos-Institucional |
| Idioma: | español |
| OAI Identifier: | oai:repositorio.unapiquitos.edu.pe:20.500.12737/11198 |
| Acceso en línea: | https://hdl.handle.net/20.500.12737/11198 |
| Access Level: | acceso abierto |
| Palabra clave: | Redes neuronales convolucionales Teléfonos móviles Melanoma cutáneo maligno https://purl.org/pe-repo/ocde/ford#2.02.04 |
| Sumario: | This research develops and validates a mobile application based on convolutional neural networks (CNN) for the diagnosis of cutaneous melanoma, in order to offer an accessible and accurate tool that facilitates early detection in non-specialized users. The problem addressed is the limited accessibility to dermatological diagnoses in populations in remote or low resource areas. To solve it, the objective is to implement an optimized CNN model in the InceptionV3 architecture in a mobile application, ensuring diagnostic accuracy and usability. The applied methodology includes the development of a CNN model adapted for mobile devices, field tests with labeled images and validation through classification metrics and user satisfaction surveys. The results reflect a 100% accuracy and concordance with the clinical diagnosis, with a Cohen's Kappa coefficient of 1.0, confirming the effectiveness of the model in identifying suspicious skin lesions. In conclusion, the app proves to be a reliable and accessible tool, whose intuitive design and diagnostic efficacy make it a potentially revolutionary option for dermatological health in areas with limited medical coverage. Future improvements could optimize the app's ability to recognize other skin conditions, strengthening its impact on public health. |
|---|