Coronavirus Response Needs Global Collaboration
Image Source: NIAID-RML
NIAID’s Rocky Mountain Laboratories (RML) produced images of the novel coronavirus on Tuesday, Feb. 11, 2020.
By: Ed Butler, VP of Corporate Development
The new coronavirus (COVID-19) outbreak hit close to home for our team at CuraCloud. We provide AI services to collaboration partners in China supporting over 100 hospitals, two of which are located in Wuhan, Hubei Province where the virus first appeared. It does not seem so far away.
We have all seen video footage from hospitals in Wuhan where a shortage of supplies has been a major concern. Less graphic, but also alarming, is the economic impact of millions of people staying away from work and avoiding public places. It is not yet clear whether the outbreak can be contained despite the strong efforts of public health authorities in China and other countries to quarantine patients with the virus.
The World Affairs Council and the Washington State China Relations Council held a public forum Feb. 13, 2020 on the response to the Coronavirus outbreak. Speakers included Dr. Ann Marie Kimball, a strategic advisor to the Rockefeller Foundation and a veteran of the SARS outbreak in 2003, Gabrielle Fitzgerald, CEO of Panorama and a veteran of the Ebola outbreak, Dennis Worsham who is managing King County’s response to COVID-19, and Jeff Hall of the Gates Foundation. The Gates Foundation is investing $100 million to support the response to this outbreak, of which $20 million is to be allocated for detection and containment, $20 million to protect the most vulnerable populations in developing countries, and $60 million to develop diagnostics, therapies, and clinical testing. The panel agreed that this investment is a “drop in the bucket” of what this outbreak will eventually cost.
Dr. Kimball contrasted the SARS epidemic of the last decade with the MERS outbreak and COVID-19. The SARS epidemic had a 10% mortality rate, and MERS had a 20% mortality rate. The current COVID-19 epidemic has spread more easily than these prior coronavirus outbreaks, but its mortality rate is about 2%. Dr. Kimball said that the recent apparent spike in cases from Wuhan last week was due to the increased use of CT scans to diagnose the disease. This was necessary in part because of the lack of enough test kits capable of definitively diagnosing the virus. Dennis Worsham related that in the U.S., there have been issues with the diagnostic kits distributed by the CDC and that they are still having to send samples to the CDC lab in Atlanta, a 10-day process. This prolongs the quarantine period for people being held who might otherwise get cleared as healthy.
It is terrifying to think of the possible spread into areas include Sub-Saharan Africa, with high population density and limited healthcare resources. We need ways to prepare the developing world for the likely spread of COVID-19 beyond the countries that are managing it today. In addition to needing more labs and more intensive care facilities, predicting where these are needed requires predictive epidemiology.
It may be possible to apply deep learning to data from mobile search engines active in Africa and other parts of the developing world to assist epidemiologists in predicting hot spots. For example, more than ten years ago, Google Flu Trends had some success in predicting influenza outbreaks as described in Nature. Even though Google subsequently shuttered this project, the potential of search data for predictive epidemiology remains a promising opportunity.
CuraCloud is committed to applying our AI expertise to help respond to this global health crisis. We are ready to collaborate with public health agencies, non-profit foundations, and other AI labs to accelerate progress.
I would like to share several articles we have been reading of the past week on COVID-19, covering the many ways AI is playing a role in prediction, detection, and treatment:
- Special Focus: 2019-nCoV, Radiological Society of North America
- CT Imaging of the 2019 Novel Coronarvirus (2019-nCoV) Pneumonia, Radiological Society of North America
- Two groups use artificial intelligence to find compounds that could fight the novel coronavirus, Chemical & Engineering News
- An AI algorithm detected the coronavirus outbreak a week before the CDC, Becker’s Hospital Review
- Coronavirus: Can AI (Artificial Intelligence) Make a Difference?, Forbes
- CT traces disease progression in novel coronavirus, Aunt Minnie