Deep Learning Algorithms in Chest Images for Pneumonia Detection

Porras F.T.; Rodriguez C.; Rodriguez D.; Lezama P.; Inquilla R.; Pomachagua Y. · Proceedings - 2022 14th IEEE International Conference on Computational Intelligence and Communication Networks, CICN 2022 (2022) · DOI: 10.1109/cicn56167.2022.10008321

Resumen

Pneumonia has become the respiratory disease that continuously causes deaths in the world; as a response to this serious problem, a literature review is performed to identify Deep Learning classification models for pneumonia detection with an accuracy higher than 95%. For the identification of the models, different architectures such as InceptionV3, MobileNet, MobileNetV2 Xception, VGG16, VGG19, DenseNet201, NasnetMobile, CNN, and LSTM were evaluated. Although they all show very acceptable accuracy indicators, which justifies their evaluation for model identification, the datasets were evaluated with chest X-ray images in different categories. As a result, it was determined that ResNet152V2 achieved an accuracy of 99.22%, which is considered one of the best models for pneumonia detection.

Instituciones: Universidad Nacional Mayor de San Marcos; Universidad Peruana de Ciencias Aplicadas; Universidad Nacional Federico Villarreal; Universidad Nacional de Cañete