High expectations for artificial intelligence (AI) have emerged amid the unprecedented crisis of the explosive global coronavirus outbreak, with hopes that AI might provide mankind with potent assistance in the fight against the pandemic.
AIs have been given full play to show off their capabilities, such as reading an amount of data far beyond the reach of human capacity to come to a conclusion in the blink of an eye. Combining this ability with AIs’ capacity for anticipating the future, software development projects for such purposes as helping medical experts make COVID-19 diagnoses and searching for therapeutic agents have been moving ahead quickly.
Prevention of Health Care System’s Collapse
The spread of new coronavirus infections has been severely taxing health care infrastructure around the world. In the United States and Europe, shortages of hospital beds and medical supplies have been deeply alarming, with increasing concern that Japan will be no exception.
It is vitally important to ensure appropriate allocation of limited personnel and resources to prevent the collapse of health care services. To realize this goal, the diagnosis of whether a person is ill with COVID-19 must be accomplished in a short span of time.
The number in Japan of facilities equipped with PCR (polymerase chain reaction) testing arrangements for viral infections, however, has been limited. Additionally, it is estimated that it will take a considerable span of time before additional PCR testing becomes available.
Increasing Testing Capacity through AI
A Beijing-based Chinese AI start-up, Infervision Technology Co., trained its AI system to recognize several thousands of coronavirus-caused pneumonia cases in a short time. Originally developed to assist medical image diagnosis based (CT) scan chest images, the system proved capable of detecting highly likely COVID-19 cases in a matter of seconds.
One of the strongest points of the AI software is its ability to examine a massive volume of images to identify characteristics and anomalies. The information produced helps medical experts understand what visual signs to look for, and improves the efficiency of diagnosis and isolation of patients for treatment.
The firm’s AI software has been adopted by several of China’s medical establishments, including in Wuhan, the initial epicenter of coronavirus outbreaks. It is receiving some use overseas as well, including by a major hospital in Rome, according to the company.
Planning for a Pandemic
Those in charge of medical care would possibly be compelled to perform triage — prioritizing patients treatments based on the likelihood of their survival — in the event hospitals or health care resources are heavily overtaxed. A research team in Wuhan has developed an AI software that is able to forecast COVID-19 patients’ chance of survival based on analysis of their blood sample.
When research team members trained the AI system using data on blood constituents, such as lymphocyte cells and enzymes of some 400 infected persons, they were reportedly successful in identifying high-risk patients with more than 90% accuracy. The system is expected to go a long way as a tool to help medical experts make decisions during an explosive outbreak.
Moreover, it is now within the realm of possibility to predict where an explosive outbreak is likely to occur, and anticipate the medical care needs, by tracing individual behavioral details and weighing it against their medical data. However, in applying such capacity, due considerations must be paid to the individual privacy rights.
In this connection, Min Sun, the chief AI scientist of Taiwan’s AI development company Appier, explained: “If planning of medical responses can be drawn up in advance based on forecasts by AI, it may become possible to rein in the expansion of COVID-19 infections such as those that could lead to the need for placing a city under a long-term lockdown. This could avoid the heavy damage to the economy caused by such a lockdown.”
AI Applied to Monitor the Virus and Search for Cures
AI can also play a major role in the speedy development of curative treatments against COVID-19. Graphen, Inc., an AI start-up headquartered in New York, has analyzed the whole genetic sequence of the new coronavirus, including variants reported from all over the world.
It was successful in developing AI software capable of keeping the virus under surveillance to track how it spreads and the mutations it spawns at various stages. It is said that understanding the genomic evolution of COVID-19 will help the development of vaccines and treatments by pharmaceutical companies and others.
Meanwhile, InSilico Medicine Hong Kong Ltd, using an AI system for drug-discovery research that the company developed on its own, has successfully identified a molecular compound that could be a good candidate in the search for medical remedies. InSilico has made its information on the molecular compound free to the public, with a view to expediting development of anti-COVID-19 medicines by researchers all over the world.
The fruits of tens of thousands of studies have been publicly released in the form of academic papers and journal articles by medical workers and researchers combating the coronavirus around the world. Although it would be impossible for humans to look through all of them, AI systems can do it. Indeed, projects have been launched with the aim of discovering new approaches to combating the coronavirus by cobbling together pieces of knowledge scattered among the volumes of research data and extracting important information from them.
There also is plenty of room for AI software to be used for practical purposes, helping medicine take a quantum leap amid the COVID-19 crisis. Appier’s AI scientist Min Sun noted: “AI can contribute to every phase of the effort to fight the coronavirus, from prevention of the disease to its diagnosis and treatment.’
He added, “Government budgetary assistance must be extended to throw support behind these efforts for utilization of AI systems for combating the virus.”
(Click here to read the original report in Japanese.)
Author: Maki Matsuda
Maki Matsuda is a staff writer of the Science and Technology News Department, The Sankei Shimbun.