Artificial Intelligence Approaches for the COVID-19 Pandemic

Abstract

Healthcare systems are now shaping up by utilising digital technology advancements that help us transform unsustainable healthcare to sustainable ones. Digital technologies like artificial intelligence, VR/AR, 3D printing, robotics, and nanotechnology play a vital role for faster and effective solutions to many diseases. COVID-19 alias coronavirus is an infectious disease that had become ubiquitous and originated from the newly uncovered coronavirus started at Wuhan in China in December 2019 and later started spreading worldwide. COVID-19 belongs to the family of RNA viruses which cause respiratory tract infection, which cause mild to fatal results. Most people who experienced COVID-19 have mild to moderate symptoms and lead to deadly results in few cases based on their health condition. Symptoms of COVID-19 are quite similar to normal flu. It is a very tough task to differentiate the actual COVID-19 victims from normal flu patients. Our study focuses on COVID-19 victim’s symptoms and clinical reports. The most time-consuming task lies in the confirmation of the disease, and sometimes to get the proper confirmation of the disease, the samples are given in two labs concurrently. Partially, this has an impact on the rise of cases day by day. Healthcare systems are now in need of decision-making techniques to control the virus from its rapid spread worldwide. Artificial intelligence plays a skilful way, like human intelligence. This paper will apply artificial intelligence techniques to analyse, prevent, and fight against the COVID-19. Records of various COVID-19-suspected patient’s data have been collected from the ICMR issued by the government district hospital, Anakapalle, Visakhapatnam. Datasets are trained based upon the collected records. Unstructured data like clinical records are applied to natural language processing (NLP) techniques, and structured data like chest X-ray images are applied to artificial neural network (ANN). Our study can help in making the correct decisions by utilising the best artificial intelligence techniques. Results obtained are simulated to identify the condition of COVID-19 patients, thereby diminishing the severity and death cases of COVID-19. Methods: Clinical records data has been given a natural language processing technique to convert human written language to machine language processing technique. It analyses the data, and chest X-ray images are given to the artificial neural network. Both the techniques are used to predict the condition of the patient based on the records. Findings: At the time of writing this paper, the total number of confirmed cases worldwide is nearly 7.82 million, and the total number of deaths occurred is 432,000. The total number of cases in India is 343,000, and the total number of deaths occurred is 9900. This count is up to mid-July, and this may increase or decrease in future depending upon the conditions. © 2022, Springer Nature Switzerland AG.