As a kid, Allyson Byrd, PhD, remembers going to her parents’ office on the weekends. Her parents, both doctors, had racks and racks of health records all in paper folders — and would have her file the folders alphabetically to keep her busy.

Artificial Intelligence (AI) and Machine Learning Fuel Research

Using EHR data in research is not a new idea. But for a long time, there was a core challenge: Hospitals and clinics use different conventions to track patient data. Because records are formatted in different ways, using EHR data was challenging because converting it into a usable format would take too much time.

Enter AI and machine learning. These tools can clean, sort and reformat the data in no time. That allows scientists like Dr. Byrd to mine these huge data sets for meaningful patterns. They can ask all kinds of questions — like is it safe for people to take two medications at the same time? Or do people with certain conditions have more side effects when taking medication?

“Some companies are even exploring using EHR data as the control arm of studies. This would allow trials to move faster because you wouldn’t need to enroll a separate control group, you could just mine existing patient data,” Dr. Byrd said.

Exploring Questions in Cancer and Autoimmune Disease

In their pilot study, Dr. Byrd and Dr. Speake will pose two questions:

Do antibiotics make checkpoint inhibitor therapy less effective for colorectal cancer?

Checkpoint inhibitors are a groundbreaking cancer treatment that harnesses a patient’s own immune system to fight off cancer. But studies in patients with melanoma and other cancers suggest that antibiotics may make checkpoint inhibitors less effective. Dr. Byrd will explore whether this holds true for colorectal cancer.

Philanthropy Fuels Advances at BRI

BRI launched the Innovation Fund to solve an ongoing challenge in biomedical research: Scientists need strong preliminary data to apply for large-scale government grants. But the govern­ment does not typically fund the research or the purchase of needed tools to collect that preliminary data.

The Innovation Fund helps bridge that gap. It helps scientists bring new tools to BRI and test ways to use them. After the pilot study, the tool becomes available for scientists across BRI. Dr. Byrd has already been exploring uses with BRI’s Bernard Khor, MD, PhD, to see if EHR data could help answer long-standing questions about effectiveness of flu vaccines in people with Down syndrome.

“This work wouldn’t be possible without the donors who support this fund, so I’d like to say thank you so much to those who have donated,” Dr. Byrd said. “These types of pilot studies help us bring in new technologies to accelerate our work, allowing us to ask novel and exciting questions. In today’s world, where grant funding is tighter and grants are more competitive, the Innovation Fund empowers us to generate the rigorous preliminary data to successfully secure those grants.”