A Modified Moth Flame Optimization Algorithm for Multi-level Classification of COVID-19 from Tuberculosis and Pneumonia Chest X Ray images using Deep learning

Abstract

Lung diseases refer to a wide range of illness conditions that affect the lungs, including pneumonia, TB, lung cancer, and numerous other respiratory issues. Recently, COVID-19 is a global pandemic infectious disease with high mortality rate which also affects lungs. The earlier lung diseases such as Pneumonia and Tuberculosis are closely related to COVID-19. All these illnesses induce severe respiratory conditions and breathing issues, which ultimately cause death. So, it is necessary to classify these diseases which helps in providing early treatment to save lives. Features plays indispensable role in classification and feature selection or optimization helps to select the most significant features. In this research, a novel optimization algorithm namely “Modified Moth Flame Optimization” (MMFO) is developed for classification of COVID-19 from Tuberculosis and Pneumonia using Chest X Ray images. The proposed MMFO address the issue of population diversity in MFO by introducing an inertia weight to balance the exploitation and global search capabilities from the perspective of diversity. The proposed MMFO algorithm is evaluated using deep learning and the experimental results are compared with state-of-the-art optimization algorithms such as Moth Flame Optimization (MFO), Grey Wolf Optimization (GWO), Crow Search Optimization Algorithm (CSA), Dragonfly Optimization Algorithm (DA), Aquila Optimizer and Whale Optimization Algorithm (WOA). Comparison results proved the superiority of the proposed MMFO algorithm.