We Can Use Artificial Intelligence Against Pandemics. Why Stop at COVID-19?

Learning from the Past

In June 2021, six months after the first identified COVID-19 outbreak  [1], the World Health Organization (WHO) issued its first global report on artificial intelligence (AI) for health. The WHO report affirmed that AI could transform healthcare for the better, emphasizing its existing applications and latent potential:

Artificial intelligence can be, and in some wealthy countries, is already being used to improve the speed and accuracy of diagnosis and screening for diseases; to assist with clinical care; strengthen health research and drug development, and support diverse public health interventions, such as disease surveillance, outbreak response, and health systems management…It could also enable resource-poor countries and rural communities [...] bridge gaps in access to health services. [2]

However, the report also advised against a rash approach to AI for health. It acknowledged that “AI holds great promise for the practice of public health” but also “that, to fully reap the benefits of AI, ethical challenges [...] must be addressed.” [3] Unbeknownst to its writers, the perfect opportunity for health innovators to capitalize on AI’s promise and address its challenges had already arrived: a pandemic of transformative proportions, COVID-19.

Artificial Intelligence v.s. COVID-19

The number of COVID-19 cases reported globally was roughly 175 million when the WHO published its 2021 report on AI for health. [4] Today, that number exceeds 577 million. [5] Although the WHO report mentions some early applications of AI for COVID-19 — e.g., contract tracing — the “smart” technology assumed far greater roles after its publication. Ultimately, COVID-19 was a kind of lab for AI for health. We can learn much from it.

Treatment Development

The Chinese technology company Baidu collaborated with Oregon State University and the University of Rochester to develop the AI-based Linearfold algorithm: an algorithm with the power to predict a virus’s secondary RNA structure far faster than prior methods. Baidu scientists used the Linearfold algorithm to predict the COVID-19 secondary RNA sequence, “reducing overall analysis time from 55 minutes to 27 seconds.” [6] Their results gave researchers a headstart in developing an mRNA vaccine for COVID-19, saving lives. 

Pandemic Forecasting

Artificial intelligence also helped forecast the spread of COVID-19. Canadian startup BlueDot identified the first known COVID-19 outbreak using machine learning  — nine days before the WHO declared the emergence of the novel coronavirus. Other AI-based forecasting tools, including HealthMap at Boston Children’s Hospital and Metabiota in San Francisco, also flagged COVID-19 (an “unusual pneumonia”) before its official recognition. [7]

These accomplishments should be taken with a grain of salt. Some individuals and human teams claim to have identified the outbreak on the same day as AI early-warning tools. [8] However, that’s not to say that AI couldn’t work with humans to improve forecasting or, perhaps with more research and data, beat humans to predict the next pandemic.

Patient Management

Finally, AI helped overwhelmed healthcare systems manage a record influx of patients amidst resource and staff shortages. Deep learning models and machine learning-based approaches helped hospitals diagnose COVID-19 and distinguish sufferers from other pneumonia cases. This AI-assisted screening accelerated the quarantine process for potentially transmissible COVID-19 patients, preventing further spread to essential healthcare workers and immunocompromised non-COVID-19 patients. [9]

Synaptiq can speak to the efficacy of AI for patient management  — not just for COVID-19. One of our solutions is a machine vision-driven application that helps prevent central line-associated bloodstream infections (CLABSIs). This type of infection is shockingly dangerous; 400,000 CLABSIs and 28,000 CLABSI-related deaths occur in the United States alone every year. [10] Our solution aims to lower those numbers by ensuring CLABSI prevention-guideline compliance. We are currently piloting a solution for hospitals to anticipate and prevent CLABSIs.

Looking to the Future

There will be pandemics after COVID-19. Just last month, the World Health Organization declared a “global emergency” in response to 16,000 cases of “monkeypox” in 75 countries. [11] It doesn’t take an oracle to predict that our future depends on mitigating pandemic threats in the present. AI is one of many tools we could use to do so. It’s not perfect — misuse is a valid concern — but AI can save lives when used ethically and responsibly.