The term “artificial intelligence” may conjure up futuristic images of flying cars or talking robots, but the way it’s used is more practical than what we see in the movies.

Dr. Eric Eskioglu
Dr. Eric Eskioglu

“Think about the sensors on your car that beep when someone is in your blind spot. Or how you interact with your Alexa and other devices that learn our behaviors over time. That’s AI at work,” said Dr. Eric Eskioglu, Novant Health executive vice president and chief medical officer.

Health care systems are also putting AI to good use. And its potential to improve the health of our communities is only expected to increase, Eskioglu said.

He explained what makes this so exciting, both now and in the future, to Healthy Headlines’ Gina DiPietro. Listen to their conversation in its entirety on the Novant Health Industry Insights podcast channel.

How is AI used in health care?

Let’s start with the question: What is data? In health care, data includes anything a patient generates throughout their lifetime – information from annual physicals, sick visits, ultrasounds, and prior hospitalizations or surgeries. This is important because we need vast amounts of data from diverse groups of people to use AI efficiently.

In some cases, patient data generates “predictive analytics,” which is another piece of the AI puzzle. Predictive analytics help us identify trends that, for example, could result in less lab work, fewer X-rays, or the ability to predict health issues before they happen. This has the potential to eliminate unnecessary costs and keep more people out of the hospital.

Currently, we’re doubling our amount of medical knowledge about every 70 days. And that’s going to increase to every 30 days. So, it’s imperative that we put this data to good use.

Will AI replace my doctor or nurse?

No. AI is not going to replace physicians and nurses. You simply can’t replace a quality like human empathy. What I do foresee is that physicians and nurses who fully embrace AI will outperform those who do not. They’ll be best equipped to care for patients with the help of things like predictive analytics.

I think we’ll see its biggest impact on primary care. Those physicians have the longest tenure with their patients. Sometimes, they have 30 or 40 years of information about someone’s medical history. That’s valuable data and AI will help us put it to good use.

How else might AI reduce costs for patients?

Here’s an example: We've been looking at uncomplicated pneumonia admissions. Most people recover and their outcomes are similar. But we’ve noticed that some physicians order more tests than others, which can drive up the cost of care.

Some physicians order the “kitchen sink” or “the Cadillac workup,” meaning they request daily labs or daily X-rays for their pneumonia patients. But other physicians might request those tests every two or three days. AI will help us reduce this variation among providers by cutting back on unnecessary testing and ultimately, reducing costs.

What else excites you about AI?

I believe that AI will be a great equalizer for health equity. We’ve been very successful working with a startup company called Viz.ai. We were one of the first in the country to adopt it when it was just at the beginning stages, and it's proven to be a hit. Patients have been true winners.

Viz.ai merges artificial intelligence with telemedicine. It allows us to detect a stroke much faster –especially for rural patients who, historically, have not had quick access to stroke care. On average, we've been able to save about 10 minutes per stroke patient, which equates to about 19 million brain cells. Quality of life has increased, while the length of hospital stays has decreased.

Preventing bias in AI algorithms

I’d like to mention one caution point. We must ensure that bias is not entered into AI algorithms. We need to make sure we have data from all populations – including people from indigenous communities, and those who are white, Black, Asian American, Latino, Chinese, etc. – because only then we can get a true picture of how to improve outcomes for patients.