Machine Learning Approaches for the Diagnosis of Pre- and Post-COVID-19 Phases

The global pandemic scenario of pre- and post-sequelae infections due to highly evolving SARS-CoV-2 mutants warrants an urgent need for its rapid and accurate analysis and prediction, which might be crucial in combating and addressing the ensuing medical challenges. Detailed descriptions of various machine learning approaches as significant tools encompassing models classifying COVID-19 disease severity, predicting the mortality of the patients, accuracy, precision, sensitivity, and specificity rates using supervised and unsupervised learning-based machine-learning approaches have been provided. A comprehensive review of modern approaches to machine learning models for diagnosis, analysis and prediction of pre- and post-effects is integrated with the effects of COVID-19. This chapter also focuses on the studies on diagnostic analyses based on machine learning approaches using available pediatric and adult COVID-19 public domain data comprising of X-ray images, CT scan images, CXR images, lung ultrasound (LUS), etc. The diagrammatic illustrations, computational models, exclusive examples, and lucid style used in this chapter will aid in an easy understanding of the complex topics discussed. This chapter will benefit those involved in healthcare and scientific brotherhood in developing COVID-19 resilience by improving quality of care and increasing patient survival.