International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 12 | Dec 2025
p-ISSN: 2395-0072
www.irjet.net
Detection of Pneumonia using Deep Neural Network technique Dr. Bhuvaneshwari K V¹, Amshu G Movva², Dhruva M³, Nithin K J⁴, Keerti M S⁵ ¹Associate Professor, Dept. of Information Science and Engineering, Bapuji Institute of Engineering and Technology, Davanagere, Karnataka, India. ²Dept. of Information Science and Engineering, Bapuji Institute of Engineering and Technology, Davanagere, Karnataka, India. ³Dept. of Information Science and Engineering, Bapuji Institute of Engineering and Technology, Davanagere, Karnataka, India. ⁴Dept. of Information Science and Engineering, Bapuji Institute of Engineering and Technology, Davanagere, Karnataka, India. ⁵Dept. of Information Science and Engineering, Bapuji Institute of Engineering and Technology, Davanagere, Karnataka, India. ---------------------------------------------------------------------***--------------------------------------------------------------------1.INTRODUCTION Abstract - Pneumonia is a serious respiratory infection that Pneumonia is one of the most critical respiratory diseases affecting millions of individuals globally, especially in developing countries with limited healthcare infrastructure. The rapid diagnosis of pneumonia is essential to prevent severe health complications and reduce mortality rates. Recent advances in artificial intelligence (AI) and deep learning (DL) have opened new possibilities for automating the diagnosis of diseases through medical imaging. Leveraging these technologies can significantly improve early detection, diagnostic accuracy, and patient outcomes, particularly in resource-limited healthcare settings.
affects the lungs and is a leading cause of death worldwide, particularly among children and the elderly. Early and accurate detection of pneumonia is essential to initiate timely treatment and reduce mortality rates. Conventional diagnostic methods, such as manual interpretation of chest X-ray images, are often time-consuming and subject to variability in human judgment, especially in areas with limited access to expert radiologists. This project aims to develop an automated system for pneumonia detection using deep learning techniques. We leverage advanced Convolutional Neural Network (CNN) architectures ResNet50 and DenseNet121 to classify chest Xray images into normal and pneumonia-affected categories. A well-curated dataset consisting of labeled X-ray images is used for training and testing the models. The system undergoes preprocessing steps to enhance image quality and standardize input data, followed by model training and evaluation using performance metrics such as accuracy, precision, recall, and F1-score. By applying deep learning techniques to medical diagnostics, the proposed system significantly enhances the accuracy and speed of pneumonia detection from chest X-rays. It serves as a dependable and efficient tool that can support healthcare professionals in making timely decisions. Moreover, this solution is especially valuable in rural or resourceconstrained settings where access to trained radiologists is limited, thus helping to bridge the gap in medical imaging diagnostics
This research focuses on developing a deep learning–based computer-aided diagnostic (CAD) system capable of detecting and classifying pneumonia from chest X-ray (CXR) images. By employing convolutional neural networks (CNNs) and the system aims to not only identify pneumonia cases but also localize infected regions for improved clinical interpretability.
1.1 Background Pneumonia is an acute respiratory infection affecting alveoli, the small air sacs in the lungs. It remains one of the leading causes of morbidity and mortality worldwide, especially among children under five and elderly populations. According to the World Health Organization (WHO), pneumonia accounted for 2.5 million deaths globally in 2019, including 672,000 children under the age of five [1]. Early diagnosis plays a crucial role in reducing fatality rates, yet accurate detection remains a major challenge in developing regions like India.
Key Words: Pneumonia Detection, Deep Neural Network, ResNet50, DenseNet121, U-Net, Medical Imaging, Deep Learning
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 1