To help enable life-saving discoveries, AI must access “underutilized, undervalued” discoveries

Of all the ways artificial intelligence is changing healthcare, one of the most promising is by revolutionizing research. These new technologies can allow all parts of the industry to understand patients and providers like never before.

It’s not just that AI-driven tools can collect more information on more patients than ever before. They can also unlock a resource that has long been largely untapped: qualitative data.

A study in the Annual Review of Public Health explains that “Public health research on chronic diseases has historically been underutilized and undervalued in qualitative methods.” This “limits the field's ability” to gain a deeper understanding of health behaviors; identify why and how remedies work or don't work; and test new theories, the study added.

There are many reasons why this happens. Quantitative data such as numbers, multiple-choice questions, and “yes or no” questions appear to be a more concrete basis for decision-making. Reading free responses to open-ended questions can be tedious. Even when researchers analyze these responses carefully, their meaning can be obscure.

But now, healthcare companies can glean insights at scale from all that unused qualitative data. The AI-driven platform with more advanced natural language processing (NLP) methods, trained on specific vocabulary of any scientific topic, can read all these responses.

These platforms can uncover trends, FAQs, areas of confusion, and more. They can provide summaries so researchers, providers, payers, and other stakeholders can gain important insights at a glance.

The best new systems combine qualitative and quantitative data, offering the best of both worlds. On its own, qualitative data is not overtly representative, while quantitative data lacks the nuance and color needed to understand the results. When AI tools unify them, it can provide three-dimensional discoveries. The tool can also suggest next steps for research, what to test or survey, which groups to focus on, and more.

This is just the beginning. The scientific community is inundated with new studies every day, many of which have overlapping topics. AI-driven platforms can collect data sets from different sources and check them for quality, repeatability, relevance, etc. These platforms can then create broader, more understandable findings to help healthcare professionals make decisions.

They can divide their conclusions based on any number of characteristics. For example, healthcare professionals can profile a specific patient and get instant feedback, highlighting the most relevant results.

Enhancing digital twins

All of these capabilities pave the way for new and better digital twins—more representative virtual representations of real people. Outside of healthcare, these are increasingly used to simulate human behavior and decision-making. Digital twins are being used in areas such as clinical trials. But these “twins” are not all equal.

The most useful information is as detailed as possible, based on a wealth of information about real people. Creating such digital twins requires a deep foundation of qualitative and quantitative data, which must be updated in real time as new information is collected about real-world patients.

With high-quality digital twins, healthcare companies open up a world of potential. They can raise issues that normally cannot be discussed due to privacy concerns. They can test multiple therapies, drugs and other therapies at the same time. These twins can also be designed to meet unique combinations of characteristics at any given time, including age, medical history, allergies, environmental factors, social determinants, and more.

None of this is meant to endanger anyone's health. As with all research, digital twins cannot definitively prove how any individual will react. Of course, real testing with real people is still necessary.

But when these AI tools are “fed” with all the information about any given drug or therapy and tasked with exploring how the digital twin responds, they can uncover important things—benefits, complications, adverse effects, risk factors, and more. They do their job more effectively when they are built using both quantitative and qualitative data.

Digital twins have a variety of use cases in healthcare. Pharmaceutical companies can learn how people feel about drugs and vaccines, as well as barriers for patients or doctors to accept new treatments, and test new marketing approaches. Providers including clinics and hospitals can use them for brand tracking. Public health agencies can use them to help design initiatives that are most likely to succeed.

In each case, the techniques used rely on the best possible collection of information. Even the most expensive and complex systems are limited by the data they can access. Therefore, as medical and healthcare organizations look for ways to move forward, qualitative data should be key. People are not numbers, and any personal descriptions, thoughts and feelings do not represent the general public. But when you put all those numbers and descriptions together, you have a much better chance of success—helping improve or even save lives.

Photo: MirageC, Getty Images

This article appeared in Medical City Influencers program. Anyone can share their thoughts on healthcare business and innovation on MedCity News through MedCity Influencers. Click here to learn how.