Comparative Analysis of Two Approaches Utilizing Different Classification Algorithms Combined with CNN for Pneumonia Prediction in Chest X-Rays

Pneumonia is a serious respiratory infection that leads to lung inflammation and the accumulation of fluid in the air sacs. Prompt and precise diagnosis plays a pivotal role in ensuring timely treatment and enhancing patient outcomes. Machine learning (ML) techniques have proven to be valuable tools for pneumonia detection, offering automated and efficient analysis of medical images, particularly chest X-rays. The objective of this research is to assess and contrast the efficacy of two disparate methodologies employing a range of classification algorithm seamlessly integrated with convolutional neural networks (CNNs) in the identification of pneumonia in chest X-rays. The dataset used in this study is a combination of the COVID-19 Radiography database and pneumonia chest X-ray images obtained from Kaggle. In the first approach, the effectiveness of CNN-SVM, CNN-LR, and CNN-KSVM is explored, while the second approach involves CNN-RF, CNN-KNN, CNN-DT, and CNN-NB. The evaluation of these different algorithms reveals that the CNN-SVM approach achieves an accuracy of 95%, making it the highest-performing method among the seven classification algorithms employed in this research.