An automated chest X-ray analysis for COVID-19, tuberculosis, and pneumonia employing ensemble learning approach

Abstract

The COVID-19 pandemic has created a demand for rapid and accurate testing to detect the coronavirus, tuberculosis (TB), and pneumonia. Chest X-ray (CXR) images have been used as a source of testing, but the manual analysis is time-consuming and prone to error. To address this issue, an ensemble-based machine-learning approach has been developed to automate the classification of COVID-19, TB, and pneumonia using CXR images. The proposed system is divided into three steps (1) pre-processing, (ii) feature extraction, and (iii) training and classification. In the pre-processing step, the proposed system converts the CXR images into grayscale and then resizes them to the targeted input shape. The histogram of oriented gradients (HOG) and local binary pattern (LBP) are used as feature descriptors to extract the salient features from the CXR image dataset. Lastly, these essential features are learned by different classifiers ,i.e., Random Forest (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Decision tree (DT), Naïve Bayes (NB), Adaptive Boosting (AdaBoost), K-Nearest Neighbor (KNN), Logistic Regression (LR), and Ensemble Model (EM), for classification of COVID-19, TB, and pneumonia. Experimental results show that the proposed strategy performed well over contemporary methods using the CXR image dataset and achieved a classification accuracy of 98%. In this study, the authors observed that integrating the outperforming Machine Learning (ML) algorithms can yield significantly better results than using them individually.