How AI can help the public health sector face future crises

Join us on November 9 to learn how to successfully innovate and achieve efficiency by upskilling and scaling citizen developers at the Low-Code/No-Code Summit. Register here.

From COVID—19 to monkeypox and intermittent polio scares, concerns around public health crises have significantly increased over the past several years.

Living in a globally connected world amidst climate change and a growing population has enabled the emergence of more frequent viruses and fostered their spread. A research study last year estimated that the probability of novel disease outbreaks will grow three-fold in the next few decades. Fortunately, there have been significant technological developments that can help minimize the impact of these global health issues. 

As health crises have increased, so too has the power and practice of artificial intelligence (AI) in support of public health. Several factors have played into this — including rapid software advances, increased connectivity, mobile communications and cloud computing — and have been spurred even faster by the urgent needs prompted by the appearance of COVID.

Consequently, the Centers for Disease Control and Prevention (CDC) have successfully used AI and machine learning (ML) technologies to fast-track their data management and analysis of different demographics and populations, allowing them to predict and track different COVID variants, vaccination distribution rates and much more. 


Low-Code/No-Code Summit

Learn how to build, scale, and govern low-code programs in a straightforward way that creates success for all this November 9. Register for your free pass today.

Register Here

Harnessing the wealth of data

The possibilities for these technologies in the public health sector don’t stop with these uses or with COVID. They can be leveraged by public health leaders to effectively inform societal health and public policy decisions and for individual health care.

Already, AI is in widespread application throughout healthcare — from adopting the right data-science platform, to moving beyond descriptive analytics, to predictive analytics for more actionable insight. It is also evident in applications ranging from predicting the clinical outcomes of COVID patients, to the discovery of new drugs to fight diseases.

One of the most significant challenges is harnessing the vast amounts of information now available from a multitude of sources to derive meaningful insights. The newer data management platforms enable public health agencies to aggregate, manage and analyze data in disparate formats. This includes unstructured data like sensor data, caseworker notes, and social media posts, most of which are not readily usable by earlier platforms.

The aggregate data can then be segmented and assessed to provide a clearer view of overall popular health. Predictive analytics based on this data can identify emerging trends, risk factors for disease and suggest allocation of limited healthcare resources. 

AI can also be used to detect the appearance and spread of new diseases.

Canadian firm Blue Dot, for instance, spotted COVID even before the World Health Organization. Their goal was to “spread knowledge faster than the diseases spread themselves.” The company designed an application using ML and AI-powered natural language processing (NLP) to monitor a myriad of online information sources to track, locate and conceptualize infectious disease spread.

They then went beyond this to predict the spread of the disease to different parts of the world, and then determine the potential consequences of it spreading. Building on these capabilities, the U.S. National Science Foundation (NSF) recently announced a grant program for research and collaborations with the goal of predicting and preventing the next infectious disease outbreak, significantly contributing to public health efforts.

Increasingly, AI and ML are being used for disease prevention and management. For example, an ML platform developed by scientists in Australia can detect signs of depression in messages people post on social media.

NLP is quickly becoming a powerful tool in public health. It is supporting analysis and data extraction from scientific literature, technical reports, health records, social media, surveys, registries and other documents to support public health functions. Among other applications, this technique was able to efficiently and accurately identify evidence of problem opioid use through the rapid analysis of vast amounts of electronic health records. NLP can also assist with disease prevention strategies through more efficient evaluation of the safety and effectiveness of interventions.

The use of these tools to scan health records is also increasing. Researchers at the University of Pennsylvania and the University of Florida, for example, recently announced that they had received a grant to use AI and ML to identify which patients are at risk of developing several inflammatory rheumatic diseases. The predictions will be derived from information already available through patient electronic health records and could greatly speed up diagnosis.

Furthermore, NLP is now being used for screening new drugs and has achieved a 97% accuracy in identifying promising drug candidates. Researchers at the University of Central Florida devised a self-attention mechanism to learn which parts of a protein interact with the drug compounds, while achieving state-of-the-art prediction performance.

Hospitals, too, are finding new uses for AI technologies. Combining AI with whole genome sequencing is leading to faster detection of infectious disease outbreaks within a hospital setting. Additionally, hospital systems are now using AI to monitor their clinical workflows for the onset of sepsis, enabling them to identify patients with sepsis or septic shock more quickly than with standard methods. Rapid detection is critical to bring down the hospital death rate. Of patients who die in hospitals, one in three has sepsis.

All told, AI technologies can provide many of the advanced tools public health organizations now need.

AI and disease diagnoses: Earlier and more accurate

All these AI tools enable disease diagnoses earlier and more accurately than ever. On an individual level, AI and ML algorithms are increasingly able to generate output for clinicians leading to better diagnoses and understandings of disease.

A recent example is a new AI model for diagnosing cognitive impairment. The researchers hope this will lead to further improvements in diagnosing Alzheimer’s disease and other neurodegenerative conditions. AI-powered chatbots are now providing valuable support and advice to patients suffering from anxiety or depression, allowing them to share their emotional issues without fear of being judged, while also providing advice. 

AI and ML advances are accelerating, mirroring the overall rate of improvement for these technologies. This ranges from drug discovery to robotics. Predictive analytics and NLP are particularly promising technologies that support evidence-informed decision making in public health. 

Because of these technologies, public health is getting better at identifying diseases and at-risk conditions in close to real-time. The hope is that this could lead to population-wide disease reduction and greater equity in healthcare access and quality.

Prasad Joshi is SVP and head at Infosys Center for Emerging Technology Solutions (iCETS).

Originally appeared on: TheSpuzz