Dr. Peter Chang — the value of unique perspectives in healthcare

With COVID-19 cases still increasing in California, it’s not hard to believe that healthcare professionals are racing to innovate new solutions to flatten the curve. However, a niche problem has generated from this public health crisis.

With limited resources and space, healthcare workers need to prioritize patients based on the severity of their condition, and getting more of these patients the resources they need before the peak of their illness, the better. Nevertheless, assessing patient status is hard, especially when it comes to predicting whether it’ll get worse.

That’s a problem researchers like Dr. Peter Chang, co-director of the Center of Artificial Intelligence in Diagnostic Medicine at the University of California Irvine is solving. 

As not only a practicing physician at the UCI Medical Center but a computer scientist, entrepreneur and professor, Chang brings a multitude of perspectives to the table, providing an effective solution to this unique problem. 

“At the end of the day, there are lots of engineers building Artificial Intelligence, but rarely any engineers that also go to a hospital and take care of patients every week. That combination doesn’t really exist. For me, I’ve always been passionate about technology and the way it intersects with healthcare,” Chang said. 

As a physician, Chang recognized the problem of quantifying severity in COVID-19 patients, and as a computer scientist, developed a program utilizing AI as a solution. He programmed a website in which healthcare workers would input pre-existing data of a patient, such as demographics and lab tests.

Then, the website would output risk percentages that determine if a patient is in need of escalated care with an accuracy score of more than 90%. Benefiting from business experiences, Chang was able to integrate the website into the UCI medical center system and it is currently in use. 

“Less than 1% of researchers can take whatever they build and actually use it. We’re pretty proud of the fact that we can do that,” he said.

Website programmed to predict whether a patient is in need of escalated care. (Image courtesy of Dr. Peter Chang)

There is one more unique perspective Chang brings: one of a professor.

As a professor at UCI, Chang runs a team of data scientists and student researchers. This summer he had around 40 students involved in programming algorithms for healthcare. While collaborating with physicians and his students, Chang is able to expedite the process of developing solutions to COVID-19. 

“In this case, I just set up a task, process data and have the students build the algorithm a doctor sees the need for. I essentially am the individual that bridges these two different worlds. All that stuff, I depend on my collaborators,” he said. 

Some applications Chang is facilitating in his lab include computer vision projects. AI is able to annotate Chest Radiographs and Chest CT scans to map out an infection in the lungs and predict a negative trend. This innovation is another solution that aids healthcare workers in prioritizing patients.

Although Chang admits that many specialists in their field can just as easily annotate medical images, for hospitals without 24-hour coverage from medical experts, these algorithms could significantly accelerate the time it takes to determine the severity of a lung infection. 

“If the algorithm helps just a few patients, that’s already a great impact,” Chang said.

Chest Radiograph in which AI annotates area of infection (Image courtesy of Dr. Peter Chang)

Chest CT Scan in which AI annotates the area of infection. (Image courtesy of Dr. Peter Chang)

According to Chang, all these unique perspectives and skills he acquired were not something he purposefully set out to achieve but rather a cumulative of interests he never gave up on. Although Chang attended medical school, he continued his passion for technology, learning computer science and AI on his own. He pursued his interests, not giving much thought to how he’d use the skills in the future. 

“Just focus on something you really enjoy at this moment. Just do that, really well. If you excel at whatever you do, that thing will become relevant and important, It’s not the other way around,” he said. 

That’s how Chang found himself, during his residency training at Columbia University, in the startup world. For him, it was only to learn how to patent his AI algorithms that he’d been working on the side. 

“I consider the experience my training wheels,” Chang said. 

Through the university, Chang learned the business side of technology: property protection, copyright and how to negotiate modest deals. 

“We were just a small team with a portfolio of various intellectual property products. That was it. It wasn’t going to be groundbreaking, but whatever I was making was interesting to me,” he said.

And that was all that mattered.

Even when Chang was given the opportunity to step away from medicine completely to focus on his startup, he turned it down, as it meant losing the unique perspective of a clinician taking care of the patient. As both the customer and the developer, Chang realized the value of his distinct position. And others recognized it as well. 

As a matter of fact, AI in healthcare at the time was not even considered a huge possibility. So when Chang reached UCI, there was no job that applied to him. 

“I invented it,” Chang said. 

Now, through teaching machine learning courses through the view of healthcare applications, Chang is training the next generation in developing different perspectives to tackle issues like COVID-19. As technology continues to enhance, the more variety of problems it could solve and it takes experts in multiple fields to see this correlation.