coronavirus rendering
PHOTO: Fusion Medical Animation

As healthcare organizations continue to deal with the fallout from the coronavirus pandemic, they can use all the help possible to make the most effective decisions. In some cases, that means they're turning to artificial intelligence (AI).

In fact, AI can improve decision-making for healthcare CIOs in three areas, according to Erick Brethenoux, research vice president at Gartner Inc. They are:

  1. Anticipation: “Predictive techniques and knowledge graphs can help with predictions and analysis.”
  2. Simulation: “Graph techniques and optimization techniques paired with potential agent techniques can help track virus spread efficiently.”
  3. Optimization: “Optimization techniques can help maximize resource allocation.”

“In the fight against COVID-19, AI ... allows predictions to be made about the spread of the virus, helps diagnose cases more quickly and accurately, measures the effectiveness of countermeasures to slow the spread, and optimizes resources, to name a few,” said Brethenoux.

Healthcare Organizations Slow to Prescribe AI

Despite the perceived benefits of artificial intelligence, most healthcare organizations are still slow to use AI in their strategic decision-making.

“Healthcare organizations are still in the early stages of adoptions,” explained Anand S. Rao, global artificial intelligence lead at PwC. “Most payers and providers have typically not used advanced analytics of AI-based techniques pre-COVID-19, like temperature screening, robotics or diagnostics. Only a small proportion of healthcare organizations — less than 20% — are well-positioned to adopt AI technologies.”

This is unfortunate, said Rao, who specializes in operationalizing AI, responsible AI and using AI for strategic decision-making.

“Healthcare organizations that embrace AI will be better prepared to not only control their current operations but can also leapfrog the competition by reimagining care delivery with remote and alternative care business models,” said Rao. “Emerging technologies — including AI, wearables, sensors, AR/VR, remote care, telemedicine — can revolutionize how we respond to the needs of patients, both during a pandemic and beyond the pandemic.”

Related Article: Will COVID-19 Be a Tipping Point for Technology?

How AI Can Support Frontline Workers

Specific to the COVID-19 pandemic, AI-based systems can help alleviate the strain on healthcare providers overwhelmed by a crushing patient load, accelerate diagnostic and reporting systems, aid in the development of new drugs to fight the disease, and help better match existing drugs to a patient’s unique genetic profile and specific systems, said Rafael Rosengarten, a board member of the Alliance for Artificial Intelligence in Healthcare (AAIH) and CEO of Genialis.

“We believe AI will also play a huge role in the next wave of this pandemic, or future outbreaks, in helping identify those most vulnerable and at-risk before it’s too late” Rosengarten explained.

Should a second wave of COVID-19 strike, one of the most valuable benefits of artificial intelligence will be the ability to plot out where and who the most vulnerable people are. This will allow key decision-makers to target interventions and lockdowns to curb the spread locally.

“We anticipate advances in Europe and Asia where we have more data available and we are seeing greater compliance with social orders and tracing. But also specific locales in the US — already some exciting work out of New York City, for example, sets the stage for interventional public health at the next serious challenge,” Rosengarten said.

Related Article: AI Transparency and the Emperor's New Clothes

5 Areas Where AI Can Improve Pandemic Responses

In reference to Gartner’s recommendations on the top benefits of AI in fighting the pandemic, Rosengarten said organizations should focus on the following:

  1. Early detection and epidemic analysis: “This is where AI can have the most immediate impact .... Wearable and connected life have made it possible to analyze broad datasets, both clinical and consumer, in real-time to enable actionable recommendations by healthcare professionals and government agencies. This blend of data and biology is particularly important for identifying biomarkers — which often are complex patterns most readily discovered by an AI system — associated with future diagnostic and predictive modeling on at-risk patient populations.”
  2. Containment: Rosengarten emphasized the importance of containment to avoid community spread and the associated health, social and economic consequences. "Containment models will derive as much from geospatial and time course data as from actual medical information,” he said.
  3. Triage and diagnosis: AI can help alleviate the lack of access to testing, a necessary step in containing the spread of the virus. He mentioned New York City-based Envisagenics as an example of a company developing AI solutions to speed diagnosis. It has developed an "AI platform that analyzes an RNA signature of the virus to complete up to 1,000 patient samples in just two hours,” said Rosengarten.
  4. Healthcare operations: “The AAIH is a member of the COVID-19 Healthcare Coalition, a private-sector led response that brings together healthcare organizations, technology firms, nonprofits, academia and startups to preserve the healthcare delivery system and help protect U.S. populations. Together, we’re coordinating our collective expertise, capabilities, and data and insights to provide data-driven, real-time insights to improve clinical outcomes,” explained Rosengarten.
  5. Vaccine research and development: Rosengarten sees a clear place for AI in the area of vaccine research and development, citing one company, Recursion Pharmaceuticals, which used its AI technology to compare "hundreds of drug candidates against cellular models of COVID infection .... to predict which drugs would have the desired therapeutic impact." Its results flagged several promising compounds as well as Remdesivir as well as showed how all other existing drugs under consideration were destined to fail. "This sort of technology-driven screening can use AI to consider a much larger search space for potential drugs and reach an accurate conclusion in a fraction of the time and cost, and without burdening human subjects,” Rosengarten said.

Most importantly, the continued spread of COVID-19 should convince healthcare organizations that have not heavily invested in AI in the past to do so now, Rosengarten argued.

“We see a huge drive in adopting AI out of simple necessity and market forces because these are the most promising solutions out there to address recalcitrant problems,” Rosengarten said. “We believe the adoption of AI solutions is poised to grow tremendously in terms of new converts as well as deep, vertical integration within organizations. This growth should be organic and based on realizing the promise of better outcomes for patients and better economics for stakeholders.”