Non-clinical Covid19 diagnosis on CT-scan, Chest X-ray, and respiratory patterns using deep-learning and signal processing

Abstract

Despite various preventative measures and therapies, the COVID-19 pandemic has exposed a number of weaknesses and vulnerabilities in global health systems, particularly in low and middle-income countries that may have less developed healthcare infrastructure and fewer resources to devote to public health. These countries have often been hit hardest by the pandemic, with higher rates of infection and death compared to more developed countries. More than 15 million deaths were reported nationwide over the first two years of the pandemic.In our thesis, we propose a novel, non-clinical method for quickly identifying COVID-19 using deep learning and signal processing techniques. This approach is based on the analysis of CT scans, chest X-rays, and respiratory patterns, and utilizes datasets containing images and audio recordings from both infected and healthy individuals. Our model is able to identify COVID-19 almost accurately using all four of these elements, making it more effec tive than other current models that only use one or two of these parameters. We believe that a non-invasive diagnostic approach could help to identify more cases of COVID-19, particularly in resource-limited settings where traditional diagnostic methods may be less accessible. As the virus continues to evolve,this method has the potential to slow the spread of the virus by enabling earlier detection and isolation of infected individuals. In addition, by providing a faster and more efficient means of diagnosis, this method can help to alleviate the burden on healthcare systems, which have been overwhelmed by the pandemic in many parts of the world.