⇓ More from ICTworks

How Artificial Intelligence Applications Can Contain Coronavirus COVID-19

By Wayan Vota on March 4, 2020

coronavirus artificial intelligence

As I write this, there are over 92,000 confirmed cases of coronavirus in 76 countries and over 3000 deaths globally. COVID-19 is spreading across continents, infecting humans at a stunning pace that’s on track to become a once-in-a-century pathogen, according to Bill Gates,.

With machine learning leading a paradigm shift in health care globally we must ask the pertinent question: “Can artificial intelligence have an impact on containing COVID-19?” No one wants a future reality where 70% of humans are infected by COVID-19 coronaviruses.

COVID-19 Containment Factors

According to the World Health Organization, a successful public health response to an outbreak of a new infection depends on four critical factors:

  • Understanding of transmissibility and risk populations;
  • Establishing the natural history of infection, including incubation period and mortality rate;
  • Identifying and characterizing the causative organism;
  • Epidemiological modelling to suggest effective prevention and control measures.

Artificial intelligence solutions can be applied to each of these factors to assist frontline health workers and supporting epidemiologist and virologist specialists in preventing, diagnosing, and treating communicable diseases like COVID-19 and other coronavirus and rhinoviruses diseases.

Artificial Intelligence and Coronavirus

Successful algorithms need copious data for training and validation to perfect its results and weed out biases, which can pose a problem when the source data is behind the Great Firewall of China or just doesn’t exists, like in poor and underserved communities worldwide.

Artificial Intelligence for COVID-19 Prevention

Researchers are trying to aggregate coronavirus data from traditional sources, like social media, newsfeeds, and airline ticketing systems to inform platforms like Healthmap, which visually represents global disease outbreaks according to location, time, and infectious disease agent.

For example, Blue Dot used natural language processing of news reports, machine learning of airline ticketing, and human epidemiologists reviews to alerts its clients of COVID-19 back in December.

More recently, researchers showed that news reports and social media reports on DXY.cn could reconstruct the progression of the COVID-19 outbreak across China and provide detailed patient-level data on its progression.

Artificial Intelligence for COVID-19 Diagnosis

Thousands of possible COVID-19 pneumonia patients have overwhelmed Chinese hospitals and coronavirus testing facilities. Software created by Beijing startup Infervision is used by 34 hospitals in China on more than 32,000 potential cases to expedite diagnosis.

Infervision’s AI software looks at lung CT scans to quickly detect lesions, and measure their volume, shape, and density to help doctors refer potential positive patients for diagnosis of the disease. The software also learns, with more scans helping the algorithm improves accuracy.

Alibaba’s research institute, Damo Academy, claims that their trained AI model uses sample data from more than 5,000 confirmed cases to diagnose coronavirus differences in CT scans with an accuracy of up to 96% within 20 seconds.

Artificial Intelligence for COVID-19 Treatment

BenevolentAI and Imperial College London used algorithms connect molecular structure data to biomedical information about relevant receptors and diseases to find potential drug targets. The software pointed to the enzyme adaptor-associated protein kinase 1 (AAK1) as a possible target for the disease. This is promising, since an associated drug, baricitinib, is already approved for rheumatoid arthritis.

Insilico Medicine recently announced that their AI algorithms designed six new molecules that could limit COVID-19 pneumonia’s ability to virally replicate in cells. These molecules are significantly different from known drugs but the reshuffled known drug motifs are predicted to be effective 3C-like protease inhibitors.

Coronavirus vs the Flu

Please note that multiple rhinoviruses and coronaviruses are already endemic in humans and there are an estimated 1 billion cases of influenza (the flu) worldwide every year that cause 646,000 deaths. COVID-19 infection size is still tiny, in comparison.

However, a CDC study of more than 44,000 confirmed cases in China found that:

  • 81% were mild with patients having mild or no pneumonia.
  • 14% were severe, for example patients having difficulty breathing
  • 5% were critical, leading to such emergencies as respiratory failure or shock

That translates into a 2% death rate overall (though 50% for those over 80 or with cardiovascular disease) versus 0.05% death rate for the flu in the USA. The other major influenza with a 2% death rate? The Spanish Flu that killed about 30 million people – 1.7% of the world population in 1918.

Filed Under: Featured, Healthcare
More About: , , , , , , ,

Written by
Wayan Vota co-founded ICTworks and is the Digital Health Director at IntraHealth International. He also co-founded Technology Salon, MERL Tech, ICTforAg, ICT4Djobs, ICT4Drinks, JadedAid, Kurante, OLPC News and a few other things. Opinions expressed here are his own and do not reflect the position of IntraHealth International or other ICTWorks sponsors.
Stay Current with ICTworksGet Regular Updates via Email

Sorry, the comment form is closed at this time.