Bibliographic citations
Meza, J., Salvador, Y. (2024). Sistema de reconocimiento de imágenes para el soporte de diagnóstico de neumonía en centros hospitalarios usando redes neuronales [Trabajo de Suficiencia Profesional, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/683772
Meza, J., Salvador, Y. Sistema de reconocimiento de imágenes para el soporte de diagnóstico de neumonía en centros hospitalarios usando redes neuronales [Trabajo de Suficiencia Profesional]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2024. http://hdl.handle.net/10757/683772
@misc{renati/1052160,
title = "Sistema de reconocimiento de imágenes para el soporte de diagnóstico de neumonía en centros hospitalarios usando redes neuronales",
author = "Salvador Romani, Yordan Brayan",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2024"
}
Pneumonia remains one of the leading causes of mortality, particularly affecting children and elderly people. Early and accurate diagnosis is crucial for effective treatment, but many healthcare centers face limitations in terms of resources and specialized personnel. This research presents an image recognition system designed to support pneumonia diagnosis in healthcare centers with limited technological infrastructure. The system employs a convolutional neural network (CNN) model, specifically the Inception V3 architecture, which was trained using a dataset of 5,215 chest X-ray images (1,341 normal and 3,874 with pneumonia) and validated with an independent set of 624 images (234 normal and 390 with pneumonia). The model demonstrated an accuracy rate of 90.71% in distinguishing pneumonia cases from normal cases. Project validation was conducted through specific success indicators, including expert approval of image analysis and algorithms, system architecture validation, and quality assurance (QA) testing, achieving a Cronbach's Alpha greater than 0.8. By integrating artificial intelligence into the diagnostic process, the proposed solution aims to improve diagnostic accuracy, reduce evaluation time, and optimize resource allocation in healthcare settings. This study significantly contributes to enhancing pneumonia diagnosis in resource-limited environments and highlights the potential of AI-driven tools to advance medical diagnostics.
This item is licensed under a Creative Commons License