How the University of South Carolina is using AI to give Americans the health care they deserve
By Abe Danaher | November 4, 2020
Homay Valafar just does not get it.
When he brings his car to the mechanic’s shop for a tune-up, it receives unique, detailed care based off its individual make, year and model. But, when he goes to the doctor’s office, he is given generic care that doesn’t take into account his unique medical history and one-of-a-kind genetic make-up.
Why, he wonders, in a country that spends over $3.6 trillion on its health care system, is more individualized care provided to his car than his body?
“A car is a relatively complex thing, but the human body is far more complex than a car,” says Valafar, the associate chair of research in the University of South Carolina’s Computer Science and Engineering Department. “So why is it that I go to a physician and they say, ‘well, you have been infected by a particular bug and I'm going to give you a generic thing. Well, here it is, take this.’? I mean, aren't you going to look at me to see who I am? See how I respond to things?”
...there’s going to be early treatment of things that wouldn’t have previously been caught, leading to early intervention rather than retroactive care.
-Homay Valafar, Computer Science and Engineering
These questions fuel Valafar’s research into the emerging field of personalized health. Alongside Amit Sheth, the founding director of South Carolina’s Artificial Intelligence Institute, Valafar and the UofSC College of Engineering and Computing have one simple goal: to use emerging technologies to improve America’s health care system so people receive care tailored to their bodies’ needs and their individual genetic make-ups.
“Care will be sooner, it will be faster, and it will be far better for you,” he says. “You are not going to be subjected to unnecessary medication. You're not going to be subjected to unnecessary treatment. And there’s going to be early treatment of things that wouldn’t have previously been caught, leading to early intervention rather than retroactive care.”
Using artificial intelligence, elderly people will receive treatment that incorporates their childhood medical history. Children will know which diseases and illnesses they are most susceptible to and be given medical guidance to combat their vulnerabilities. Patients with chronic illnesses will be empowered and provided with knowledge that allows for better self-management. Generic care will be replaced by safer, more effective, personal treatment.
Valafar gives an example of what these personalized health care conversations could be like. Imitating a physician, he says, “Hey look, based on your genetic makeup you're not susceptible to weight gain. So, if you want to visit your favorite fast food joint every so often, your body is going to tolerate it more than the next person. However, in turn, your respiratory system is far more susceptible than the rest, meaning that if you're exposed to secondhand smoking, you’re going to develop lung cancer.”
He believes widespread personalized health care is only a few years away.
Available technology has already made it possible to electronically track a person’s health data across the span of their life. Now, the College of Engineering and Computing plans to use advanced artificial intelligence techniques to interpret the data so that all health care providers can offer personalized insights to their patients.
Currently, Valafar is working with vascular surgeons at Prisma Health to use artificial intelligence to identify critical artery blockages in CAT scans. He is also collaborating with researchers in the UofSC College of Nursing to detect medication adherence using AI. Sheth is using his augmented personalized health system, named kHealth, to help patients with chronic illnesses — like asthma — self-monitor, self-appraise and self-manage. Together with the college, they are working toward a comprehensive vision to improve the entire health care system from top to bottom.
“We don't aim to replace the physician. We aim to assist the physician with the pertinent information,” Valafar says. “To gather information, highlight the pertinent information, bring it to the attention of the clinician and say, ‘Hey, when this person was 10 years old, he had a hernia. Could this possibly have anything to do with his stomachache at the age of 75?’ Artificial intelligence can assist physicians in condensing all this information.”