RSUSCI-2021 & RSUSOC-2021

IN21-029 Automatic Pneumonia Screening from X-Ray Image Using Resnet-50 Convolutional Neural Network

Presenter: Thaman Toobunterng
College of Biomedical Engineering, College of Biomedical Engineering, Rangsit University

Abstract

Coronavirus disease (COVID-19) is a current global severe health concern that causes a high death risk due to significant alveolar injury and gradual respiratory failure. There are increasing numbers of COVID-19 patients, and they are difficult to be diagnosed and classified. Thanks to one of its primary symptoms, which is Pneumonia, the diagnosis typically involves chest x-ray imaging interpreted by a specialized medical doctor or a radiologist. In this study, artificial intelligence based on Convolutional Neural Network has been developed to identify and classify patients with Pneumonia and assist medical practitioners and doctors in diagnosis as a second opinion. This deep learning software aims for the rapid diagnosis of an X-ray image. Here, the deep learning model was trained with more than 3,000 chest X-ray images with Pneumonia and healthy cases. The Resnet-50 convolutional neural network can successfully perform the classification task of Pneumonia with a classification accuracy of 99.66%.

Citation format:

Buasomboon, T., Toobunterng, T., Thanawasumongkol, T., Thongpance, N., Suvarnaphaet, P., & Pechprasarn, S.. (2021). Automatic Pneumonia Screening from X-Ray Image Using Resnet-50 Convolutional Neural Network. Proceeding in RSU International Research Conference, April 30, 2021. Pathum Thani, Thailand.