The big impact of micro-grants
Grishma Reddy (right) and Behnaz Eslami (left).
By Daniel P. Smith
For the last six years, Samie Tootooni and his research team at 黑料门University Chicago’s Parkinson School of Health Sciences and Public Health have been focused on leveraging artificial intelligence (AI) to streamline health care decision making.
Involving new technologies and large data sets, the work is challenging and complex but also invigorating. Unlocking appropriate and scalable solutions, Tootooni says, can tangibly impact lives, spurring faster, more informed care, reducing errors, heightening efficiencies, and lowering costs.
Today, an upstart Parkinson School program is propelling one of the Tootooni lab’s most ambitious projects and providing one of his student researchers a particularly dynamic experience.
Launched at the beginning of 2024, the Parkinson micro-grants program for student researchers enabled Tootooni to fully devote one of his lab’s researchers, computer science PhD student Behnaz Eslami, to an exciting project he calls “concept extraction from clinical notes.” Leveraging large-language models, the effort synthesizes information from patient notes and imagery and then highlights relevant health issues for clinicians.
“With the algorithms and methods we use, we're able to help clinicians make faster, better decisions,” Eslami says.
A rich research experience
A student in Loyola’s MD/MPH program, Grishma Reddy was among the micro-grant program’s inaugural 11 MRAs. She worked with Sparkle Springfield, assistant professor of public health sciences, on a project exploring social determinants of health as potential barriers to medication adherence among lung and heart transplant patients.
As Springfield’s MRA, Reddy helped with manuscript production during the second half of her MPH year. Alongside Springfield, Reddy interpreted data provided by biostatisticians, conducted a literature review of more than 80 references, and penned the manuscript’s first preliminary drafts.
“I value interdisciplinary, collaborative work and this project put that into action with the ultimate goal of improving patient outcomes by better understanding the patient experience,” Reddy says.
With the support of the micro-grant research program, Reddy was able to forge a relationship with an “incredible faculty member,” gain research experience beyond the clinical realm, and learn firsthand how a research project moves from the lab into the public sphere.
“As an undergraduate, I often participated at the beginning phases of research projects, but was never able to see a project through,” Reddy says. “In this case, I joined the project midstream and was able to understand the latter stage of research and the publishing process.”
Now into her MD studies, Reddy has continued working with Springfield outside of the micro-grants program. In addition to recently submitting a manuscript for potential publication, Springfield and Reddy have also begun exploring the same data with a focus on diet.
“It’s important to ask new questions, especially those rarely discussed in a clinical setting,” Reddy says, adding that her work with Springfield has further inspired her to bring a public health perspective into clinical care. “When you have an interdisciplinary collaboration that integrates social determinants of health, you have a better shot at improving patient care because you're identifying risks, tailoring patient and community-level interventions, and, ultimately, improving survival, quality of life, and equity in patient outcomes.”
As for Eslami, she calls it “inspiring” to work alongside Tootooni on a project with the promise of advancing health care and improving lives. By leveraging AI to structure notes for a clinician, Eslami is contributing to a more responsive health care system.
“Our main goal is to be an assistant next to the physician empowering them to make more informed decisions,” Eslami says. “If we can give better tools, then we can help improve outcomes, which is why it’s so incredibly exciting to be involved in a project like this.”