Pneumonia Detection using Deep Learning

Abstract

Pneumonia is a serious respiratory infection that affects the lungs, often caused by bacteria or environmental factors, leading to fluid accumulation in the alveoli.Accurate diagnosis of pneumonia is crucial for effective treatment, and it typically involves methods such as physical exams, chest X-rays, ultrasounds, and lung biopsies. However, misdiagnosis and delayed treatment can lead to severe complications. This research explores the use of deep learning techniques, particularly Convolutional Neural Networks (CNNs), to enhance pneumonia detection. By analyzing chest X-ray images, the study develops a CNN model that classifies patients as either suffering from pneumonia or not. A dataset consisting of 20,000 high-resolution chest X-ray images (224x224) was used, with a batch size of 32 for model training. The CNN model achieved a 95% accuracy rate, demonstrating its potential for diagnosing pneumonia effectively. The model is capable of distinguishing between various types of pneumonia, including bacterial, viral, and COVID-19, solely based on chest X-ray images, showcasing its accuracy in medical diagnostics.

Introduction

Pneumonia is a major inflammatory lung disease and a leading cause of death worldwide, particularly affecting children under five and adults over 65. In 2015, India recorded nearly 297,000 child deaths from pneumonia. Despite advances in diagnostic imaging like CT and MRI, chest X-rays remain the most accessible but challenging diagnostic tool due to symptom overlap with other diseases and slow traditional methods.

To improve early detection, this study proposes a deep learning approach using a VGG-based Convolutional Neural Network (CNN) to automatically detect pneumonia from chest X-rays. The model achieved a high accuracy of 96.07% and an AUC of 0.9911 on a dataset of 5,863 pediatric chest X-rays from Kaggle.

Pneumonia is classified by cause (infectious: bacterial, viral, fungal; non-infectious: immune or radiation-related) and by origin (community-acquired, hospital-acquired, ventilator-associated), with hospital-acquired pneumonia being harder to treat due to antibiotic resistance.

Related work highlights challenges such as limited COVID-19 X-ray datasets and uses transfer learning and ensemble models to improve detection accuracy. Deep CNN architectures like ResNet50 and ensembles of RetinaNet and Mask R-CNN have been used to classify and localize pneumonia, employing advanced techniques like Feature Pyramid Networks and residual connections to enhance performance.

The study details data preprocessing methods including rescaling, shearing, zooming, and horizontal flipping to augment training images. The CNN architecture uses convolutional and max-pooling layers with ReLU activation, followed by fully connected layers with softmax or sigmoid activations for classification.

Conclusion

Pneumonia is an inflammatory condition of the lung death due to pulmonary infections in 2013. In Europe, pneumococcal disease affected 35% of hospitalized patients, with a global prevalence of 27.3%. A report from the Johns Hopkins Bloomberg School of Public Health indicated that India had the highest mortality rate from pneumonia, accounting for nearly 297,000 child deaths under five in 2015. Furthermore, pneumonia remains the leading cause of death in children under five. Although pneumonia-related mortality declines with age, its incidence increases in older adults, particularly those over sixty-five. The high mortality rate among infants has spurred calls for improved detection methods.

References

[1] VandeciaFernandes et al., “Bayesian convolutional neural network estimation for pediatric pneumonia detection and diagnosis”, Computer Methods and Programs in Biomedicine, Elsevier, 2021 [2] Hongen Lu et al., “Transfer Learning from Pneumonia to COVID-19”, Asia-Pacific on Computer Science and Data Engineering (CSDE), 2020 IEEE [3] Sammy V. Militante et al., “Pneumonia and COVID- 19 Detection using Convolutional Neural Networks”, 2020 the third International on Vocational Education and Electrical Engineering (ICVEE), IEEE, 2021 [4] Nanette V. Dionisio et al., “Pneumonia Detection through Adaptive Deep Learning Models of Convolutional Neural Networks”, 2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC 2020), 8 August 2020 [5] Md. JahidHasan et al., “Deep Learning-based Detection and Segmentation of COVID-19 & Pneumonia on Chest X-ray Image”, 2021 International Information and Communication Technology for Sustainable Development (ICICT4SD), 27-28 February 2021[6]https://www.kaggle.com/paultimothymooney/chest- xray-pneumonia [6] LeCun, Y.; Boser, B.; Denker, J.S.; Henderson, D.; Howard, R.E.; Hubbard, W.; Jackel, L.D. Backpropagation applied to handwritten zip code recognition. Neural Comput. 1989, 1, 541–551. [7] Krizhevsky, A.; Sutskever, I.; Hinton, G.E. Imagenet classification with deep convolutional neural networks. Adv. Neural Inf. Process. Syst. 2012, 25, 1097–1105. [8] Simonyan, K.; Zisserman, A. Very deep convolutional networks for large-scale image recognition. arXiv 2014, arXiv:1409.1556 [9] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam,D. Parikh and D. Batra, \"Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization,\" 2017 IEEE International Conference on Computer Vision (ICCV), Venice, 2017, pp. 618-626. [10] L. Wang and A. Wong, \"COVID-Net: A tailored deep convolutional neural network design for detection of COVID-19 cases from chest radiography images,\" arXiv:2003.09871, 2020.

Copyright

Copyright © 2025 Ms. Sneha Vanjari, Tejas Abhyankar, Hrushikesh Karandikar, Rutvik Shaiwale, Siddarth Mandke . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.