How AI Can Improve Health Outcomes in the Next Pandemic

By Erico Teixeira Erico Teixeira has been verified by Muck Rack's editorial team
Published on June 6, 2023

To minimize the human collateral damage of future global health emergencies requires open access to anonymized patient data, expanse of  multi-ethnic data sets, and broadened global collaboration to act quickly. 

The Covid-19 pandemic from 2020-2022 spurred an unprecedented push for some of the world’s brightest minds to come together to save as many lives as possible. A variety of scientists played a key role in developing and trialing highly effective vaccines quickly.

One prestigious institution – The Alan Turing Institute in London – found significant room for improvement in how AI and data science can help improve health outcomes in the next pandemic, per a report it published last year. 

Three key findings on how to improve outcomes in future emergencies include: 

  1. The standardization of global access to robust and timely data
  2. Increasing the available data on underrepresented minority populations
  3. Broadening communication channels to share findings and act on them quickly

If the global STEM community can overcome these critical challenges, the health outcomes of people all over the world can be improved in the next pandemic by harnessing AI and data science to speed drug discovery, analyze people’s behaviors in public to reduce exposure, and even leverage ever-advancing automation and robots to help care for people infected with a new deadly pathogen in the future. 

One example of this integration between AI and drug design is in the prediction of chemical properties of proteins, such as free energy of binding for protein–protein complexes covered in a paper published in the Physical Chemistry Chemical Physics journal in March. This new breakthrough uses a novel Artificial Neural Network (ANN) model, and it has a variety of applications in molecular and chemical biology, materials science, and biotechnology.  

Of course, machine learning or deep learning is not the final answer; it’s the middle of the answer. It takes a “global village” of STEM experts to help to create and deploy effective vaccines, treatments, or ways to prevent the rapid spread of a new virus or other pathogen. 

The Need for More Open Access to Anonymized Patient Data 

In order to gain access to enough fresh, timely data to run through super-fast computers such as those being used in the Cleveland Clinic’s Discovery Accelerator in collaboration with IBM,  it’s critical for countries around the globe to come together and collaborate on new international policies to allow anonymized patient data to be more accessible, and in real time. 

For example, the ACT-Accelerator partnership, launched by the W.H.O. and partners, led to the fastest, most coordinated, and successful global effort in history to develop tools to fight a disease. It led to advances in RD&I by academia, the private sector, and government initiatives. 

An academic paper published in Nature Medicine in October 2022 laid the groundwork for a new policy framework to increase the speed of data-enabled responses to pandemics based on places that disproportionately contributed to policy-relevant insights, including Iceland, Israel, Qatar, Scotland, and Taiwan. 

For example, based on the paper’s findings, there’s a need to “access disparate data – including electronic health records, travel and other health-related data – ideally on every person, in as close to real-time as possible.” There’s also a need for more trained staff who are familiar with these datasets who can “pace, check, clean, link, analyze and help to visualize data for policy audiences and others,” in addition to heightened transparency and international cooperation. 

Despite concerns about data privacy, a majority of U.S. adults (71%) are “open to sharing de-identified health data for various reasons, including to improve their own or other patients’ healthcare, for research, to improve hospital services or to advance equity,” according to a recent survey by Q-Centrix, a clinical data management company. 

Addressing Data Inequality and Exclusion 

In Brazil, where I live and work, we have great examples of how AI connects with healthcare. CyberLabs, the biggest AI Lab in Brazil, rolled out advanced technologies to detect Brazilians who were infected with Covid-19 by analyzing their coughs. It also spearheaded an AI-powered monitoring system to detect people who were gathering too close together in the streets of Rio de Janeiro using networked cameras installed during the Olympics hosted in the beachside city. 

However, when it comes to AI and data, you need to be aware of regional biases and peculiarities. For example, Brazil is home to one of the world’s most diverse populations – formed by its indigenous people, European immigration, and the African slave trade. For many years, the world’s seventh most populous country was “out of the loop” for data inclusion. 

In terms of customizing AI for Brazilians, “to do facial recognition properly in Brazil, for example, you need to train algorithms with Brazilian faces. Otherwise, you are going to end up with racially biased AI. If you want to do voice recognition and voice synthesization, you need to train the AI with different Brazilians’ voices,” said CyberLabs CEO/Founder Marcelo Sales during a BayBrazil Conference in Silicon Valley in 2020 during the midst of the Covid-19 pandemic. 

CyberLabs’ work is just one example that underscores the need for a more diverse, inclusive AI culture, as well as deliberate steps toward the decolonization of data at the global level.

Broadening Global Channels to Communicate Key Findings and Act Fast

“Nothing in science has any value if it is not communicated,” said Anne Roe, the noted twentieth-century American psychologist and writer, nearly 70 years ago. With some top U.K. scientists now warning that “next pandemic is coming, and we’re not ready for it,” per a news story in The  Independent published in mid-April – it’s critical for the world’s STEM community to be ready to work together to quickly monitor for future threats and communicate their key findings. 

As we pass the three-year anniversary of when the Covid-19 pandemic began, some of the world’s leading scientists are cautioning that the next unknown virus that could cause the next pandemic may require an entirely different approach and set of tools to overcome. 

Dubbed “Disease X” by the World Health Organization, this future unknown threat has led to a number of simulated emergency meetings of the W.H.O. in order to be better prepared, according to a recent opinion piece by Dr. Tom Inglesby, the director of the John Hopkins Center for Health Security and an expert in pandemic preparedness, published in The New York Times. 

Dr. Inglesby and others argue that there’s an urgent need to prepare now, be ready to communicate key findings between countries, agree on how to react collectively to minimize the spread of new infections, and the ability for us all to act on them quickly and decisively is critical to reducing the collateral impact on human society for when, not if, the next pandemic strikes. 

“We are likely to face similar or worse pandemic threats in the future. We need to use the time we have now to make big preparedness changes to protect us from challenges that could arise again without warning,” writes Dr. Inglesby.

By Erico Teixeira Erico Teixeira has been verified by Muck Rack's editorial team

Erico Souza Teixira is a professor of Computer Science, Design, and Software Engineering at the CESAR School and Innovation Center located in Recife, Brazil.

Read more

More GD News