5 ways AI is used to fight COVID-19
We explain how AI has become one of humans\u2019 ace cards in handling the COVID-19 crisis
From its epicenter in China, the novel virus called COVID-19 has spread to at least 160 countries worldwide, across a three month period from January 2020 till date. COVID-19 has, in a short period of time, emerged as one of the biggest challenges to face the 21st century world.
As the research detail started to emerge, more and more data related to the virus is becoming available, triggering the possibility to use AI to fight this unprecedented crisis.
In this article, we are going to explain how AI has become one of humans' ace cards in handling the COVID-19 crisis.
Artificial Intelligence can be used for conducting a diagnosis on patients infected with the new coronavirus. For example, the Chinese technology giant Alibaba recently developed an AI system for diagnosing the COVID-19 virus.
In a report published by Nikkei\u2019s Asian Review, Alibaba claims its new AI system can detect coronavirus on a CT scans of patient's chests with a 96% accuracy against viral pneumonia cases. According to Alibaba, it only takes 20 seconds for the AI model to make a determination, whereas with humans it generally takes about 15 minutes to diagnose the illness as there can be upwards of 300 images to evaluate.
The AI system was trained on images and data from 5,000 confirmed coronavirus cases and has already been tested in hospitals throughout China. According to the Asian Review's report, at least 100 healthcare facilities are currently employing Alibaba's AI model to diagnose COVID-19 on their patients.
Alibaba is not alone in trying to fight the coronavirus with AI. Some companies are selling their tools, others have released free online versions, and various groups are organizing large crowdsources repositories of medical images in order to help in generating new AI algorithms. Recently, RADLogics published a research, validating its AI-powered system trained on multiple international datasets. After a chest scan, the system can give immediate alerts if a patient needs to be seen by a healthcare professional. The tool also tracks a patient's progress by providing a numerical "Corona score" - a measurement of disease severity, which can be used to quantify the disease over time. The software is currently being deployed in China, Russia, and Italy.
A new study of American and Chinese researchers announced that they have developed a tool using Artificial Intelligence to predict which coronavirus patients will develop serious lung complications. Once deployed, the tool could allow doctors to treat certain patients as a priority while the Health systems in many countries around the world are running out of steam to cope with the spread of the virus.
The tool has uncovered several indicators that strongly predict which patient may develop acute respiratory distress syndrome (ARDS), a side effect complication of COVID-19 that results in filling human lungs with fluid and is responsible for the death of 50% of people who developed it. This AI tool made it possible to predict the risk of ARDS patients with an accuracy of up to 80%.
Artificial Intelligence is already being used by dermatologists to predict which patients are at risk of developing skin cancer. In the case of COVID-19, a disease still poorly understood, the tool can lead healthcare practitioners in the right direction of knowing which patients to treat first if hospitals are overloaded with patients.
Prediction and management at large scale
The Belgian based startup Kantify has been able to rapidly develop a machine learning model that can very accurately predict the spread of COVID-19. This model integrates many variables, which makes it very different from traditional epidemiologic models.
Thanks to Machine Learning, the company is not only able to predict highly accurately the spread of the virus throughout the world, but to determine what are the exact parameters that influence the spread. This can be used as a decision-making system to influence policies and measures to mitigate the spread of the virus, while exactly knowing what can be the impact of the model.
This model can also be used to understand how global supply chains are and may be impacted by COVID-19.
Taiwan being 130km off the coast of mainland China, was the country expected to have the second-highest number of cases of COVID-19 due to its proximity and the number of flights between China. But so far Taiwan reports that it has largely mitigated the spread of the pathogen. The country owns its success largely to the emergency implementation of big data analytics and new technologies, according to a report in the Journal of the American Medical Association (JAMA). According to the report, Taiwan officials "did a very detailed mapping of who got the virus from the beginning of the viral outbreak, and were able to stop a lot of transmissions early". Notably, officials integrated Taiwan's national healthcare insurance database with its immigration and customs database. This enabled the government to track travel histories and symptoms of its citizens and give access to the platform to all hospitals, clinics, and pharmacies for each patient. All the action was been oversaw by "data analysts and reporters, which can sometimes host up to a hundred people 24/7".
It is very likely that like the rest of the countries of the world, the number of officially confirmed cases in Taiwan is likely far higher than the true number on the ground, since there are people who have the disease and don't know it, or have such mild symptoms that they don't seek care or testing. Nevertheless, the country had managed to keep an extremely low number of confirmed cases as opposed to what is has been expected, by quickly responding to the virus thread by using big data and implementing new technologies.
The novel virus has been circulating among humans for barely three months, but several biotech firms have already created drugs that target the COVID-19 disease. One of the secret weapons for fast response is Artificial Intelligence.
The researchers around the world applauded for the quick response of Chinese scientists in decoding the genetic sequence of the virus, dubbed SARS-CoV-2, and posting the results in a public database on January 10. It took one biotech firm in San Diego, Inovio Pharmaceuticals several hours over the course of that weekend to generate a vaccine candidate based on the genetic blueprint received from the Chinese researchers. Innovie plugged the viral sequence into its machine learning system, dubbed SynCon, which took just three hours to generate a "fully designed" DNA-based vaccine for prevention of infection with the COVID-19 virus, called INO-4800. The drug is currently in preclinical trial, with animal trials starting soon.
Moderna Therapeutics took a similar approach with its modified messenger RNA (mRNA)-based vaccine for COVID-19, which is also in a clinical trial. Once the genetic blueprint was made public, it took researchers from Moderna just two days to finalize the sequence for their vaccine, and twenty-three days later, Moderna had the mRNA-1273 vaccine ready for a clinical trial.
Drug development is a traditionally very slow process based on trial and error. Getting drugs to human clinical trials can often take months, if not years. However, with COVID-19 threatening many human lives across the world, the researchers are teaming up with Artificial Intelligence to discover a treatment that could stop the virus outbreak.
If you are curious about AI and the ways your organization can leverage this technology to make smarter and data-driven decisions, book a short introductory slot, or simply send us your questions and our team will answer them shortly.
Last but not least, make sure you subscribe to the Kantify newsletter where we regularly share insights about applied machine learning.