IMG-LOGO

Artificial Intelligence to tame Coronavirus

by Anuj Kumar Sirohi - 24 July, 2020, 12:00 573 Views 0 Comment

“It is during our darkest moments that we must focus to see the light.” – Aristotle Onassis

From the sixth century’s Justinian Plague to the Spanish flu of the last century (1918-19) humanity has faced several devastating pandemics. Before the first identification of a disease caused by a microorganism named ‘virus’ (‘contagium vivum fluidum’ named in Latin) in 1898 by a Dutch microbiologist, Martinus Beijerinck, all infectious diseases were supposed to be caused by invisible bacterias. Prior to the discovery of modern antibiotics in 1928 by Alexander Fleming, doctors used to give a different treatment namely ‘phage therapy’ to treat bacterial infections. In phage therapy, a special virus harmless to humans but which kills the bacteria is being used. In the 1920s, several institutes in Europe pioneered the use of this therapy to treat bacterial infections in humans. The advent of antibiotics brought a paradigm shift in the treatment process of bacterial infection, although phage therapy is still used by doctors to treat certain multidrug-resistant bacterial infections.

Further, recent advancements in the field of biotechnology, viruses can be used to cure some non-infectious ailments. In virotherapy, a virus can be used for medication to treat a disease by reprogramming it through genetic manipulations to act as a therapeutic agent. In 2015, the FDA (US) had approved a genetically modified herpes virus, Talimogene laherparepvec, to treat inoperable melanoma, a type of skin cancer. Similarly, in 2017, a virus vector-based novel gene therapy, Luxturna, was approved for treating retinal dystrophy in adults. Also, in 2019, the FDA approved another novel virus vector-based gene therapy,  Zolgensma, to treat a genetic disorder, spinal muscular atrophy in children. Several other virus vector-based treatments for genetic disorders, cancer and immunodeficiencies have been approved.

However, despite these spectacular successes in medical science, the mysterious nature, and the astounding speed and scale of replication and proliferation, viral infection outbreaks have always possessed challenges and endangered humans’ lives. But, every time humankind faced a new challenge due to novel infection, it had invented new ways to overcome it successfully. Scientists had been relentlessly putting efforts to develop vaccines as soon as possible to prevent infection or medication to mitigate its effects to reduce the loss of life. Except for few serendipity, generically developing a vaccine or cure for any disease is a time taking arduous endeavor. However, the recent emergence of cutting edge disruptive digital technologies has brought a revolutionary change in drug-development. It is indisputably evident that current and future drug and vaccine development will be powered by machine (deep) learning and artificial intelligence (AI).                    

Recently scientists have used artificial intelligence to develop a novel potential antimicrobial compound. In February 2020, an innovative deep-learning model was developed by scientists at Massachusetts Institute of Technology, USA for designing antibiotics for killing multidrug-resistant bacterias.         

Currently, though the coronavirus disease 2019 (COVID-19) pandemic outbreak started in Wuhan China but has been spread across the globe within weeks. The COVID-19  disease caused by a novel coronavirus SARS-CoV-2 has endangered the lives across the continents and brought complete standstill of all human activities. Across the world, countries have applied strict infection-prevention public-health measures such as social distancing to contain the pandemic. But, unlike 2003’s Severe Acute Respiratory Syndrome (SARS) epidemics, in 2020, our classical response armory to tame outbreak is additionally equipped with AI-based tools. Machine learning-based AI-tools are being applied in all public-health measures from monitoring, surveillance, diagnosis to drug development to contain the pandemic outbreak and mitigate its ill effects.

In this direction, BlueDot, a Canadian AI company was first to forecast pneumonia like pandemic (COVID-19) outbreak in Wuhan and it also helped in identifying new hotspots so health agencies can take preventive steps in advance. China and Russia have deployed thermal imaging enabled AI-based multisensory facial recognition tools for contactless identification of people with potential symptoms at screening points. Machine (deep) learning based AI tools have been developed for diagnosis of COVID-19 based on different kinds of sample data. Researchers have made impressive progress by genome sequencing and AI based structure prediction of coronavirus proteins. AI is also playing a crucial role in developing vaccines and drugs to cure COVID-19.

Since China, an AI technology hub, is the epicenter of the COVID-19 outbreak is playing a leading role in the application of AI to tackle outbreaks. The LinkingMed, a Beijing-based data analysis company and Chinese search engine Baidu’s PaddleSeg, developed ‘inferVISION’ to diagnose COVID-19 within seconds with high accuracy from CT scan which substantially reduces the requirement of confirmatory­test based on Reverse Transcrip­tase Polymerase Chain Reac­tion (RT­PCR). Similarly, Baidu’s AI team has developed a tool called LinearFold to reduce significant time in the prediction of the secondary structure of virus RNA. Also, Insilico Medicine, a Hong Kong based company has developed potential drug molecules based on AI techniques generative adversarial networks (GANs). Shenzhen based Pudu Technology have developed robots to help medical staff to serve patients without coming in contact. China is also using drones to monitor, surveillance the outbreak and to disinfect regions.  

DeepMind, Google’s subsidiary, is applying its AI-system namely AlphaFold for predicting structures of different proteins with coronavirus (SARS-CoV-2). The UK based BenevolentAI has come up with a protein kinase based possible drug called Baricitinib with the help of AI-based software tools. Another Oxford-based AI-assisted drug discovery company, Exscientia is working on similar projects. Andrew Hopkins, chief executive officer of Exscientia Ltd. anticipated that new treatments from conception to clinical testing can be materialised in short periods in the coming decade with the help of AI. Several other companies including from Israel have catapulted their AI resources to make strides in developing drugs for coronavirus.

The sobering conclusion is that the whole world is still grappling with COVID-19, more than six lakh people have died. It is premature to aggrandise the AI’s importance in taming down the pandemic. However, the AI community can keep its chutzpah and draw some inspiration from frontline health workers and doctors to augment their efforts in fighting this calamity.

Anuj Kumar Sirohi
Anuj Kumar Sirohi
Author is a doctoral candidate at School of Computational and Integrative Sciences, Jawaharlal Nehru University, New Delhi, India. His research interests include the application and philosophy of artificial intelligence.

1 reply on “Artificial Intelligence to tame Coronavirus”

Avatar

This is a very poorly formed article. It sounds more like good carpentry. Though a bit informative but fails to even use the jargons correctly. There is nothing new in this article except a collage of newspaper headlines. Grammatical error and typos have plagued this article. Its clickbait is very nice, headlines goes like it more about the details of the application of AI but it feignedly describes application of AI in CoVID.



Leave a Reply

Your email address will not be published. Required fields are marked *