Pneumonia Detection in Chest X-Ray Images by using Resnet-50 Deep Learning Algorithm

• Published in: 2023 Third International Conference on Artificial Intelligence and Smart Energy (ICAIS) pp. 1078 - 1084
• Main Authors: Kadali, Vasavi, Pudi, Bhavani Shankar, Shaik, Khaleel Ahmed, Janjam, Anuhya, Javvadi, Jagadeesh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 02.02.2023
• Subjects: Chest X-ray images; Computational modeling; Computer-aided diagnostics; Convolution Neural Network (CNN); Convolutional neural networks; Deep learning; Neural networks; Pneumonia; Proposals; Pulmonary diseases; Resnet-50; Transfer learning; X-ray imaging
• DOI: 10.1109/ICAIS56108.2023.10073748

A number of different viral infections can result in pneumonia, a serious lung infection that can be fatal. Due to pneumonia's resemblance to other lung diseases, diagnosing and treating it from chest X-ray pictures can be challenging. As a result, significant levels of accuracy cannot be achieved by the current approaches for forecasting pneumonia. In order to make the diagnosis of pneumonia on chest X-ray pictures simpler, our work introduces Transfer learning characterization of pneumonia, which is computer-aided. Convolutional Neural Network (CNN) models, these are pre-trained CNN models used in our proposal rather than freshly trained CNN models, have lately been used to improve the performance of numerous medical activities. This study considers using RESNET50 model, which has been pre-trained on the ImageNet database. Using fine-tuning, on the chest X-ray dataset these models are being trained. By merging the features that were derived from these 3 models throughout the experimental phase, the findings are finally obtained. With an accuracy of 94.01% on the testing phase, the suggested Transfer Learning strategy beats other current state-of-the-art methodologies.