UVA Engineering Ph.D. student helps cardiovascular surgeon create a lifeline to better treatment for heart failure patients
Josephine Lamp came to the University of Virginia School of Engineering to earn her Ph.D., armed with a passion for using technology to solve complex medical problems. It was that same drive for solutions that led Lamp to computer science as a teenager.
“I was a dancer in high school, and I wanted to find a way to build a light-up tutu that dynamically changed colors to music,” Lamp said. Her computer science teacher taught her how to write code and solder to create the engineering wonder.
“The tutu was pretty cool when it was finished, but I think the biggest thing I took away was how awesome computer science was as a tool,” she said.
Now she is using computer science as the tool to help heart failure patients get the most advanced care.
After receiving her bachelor’s in biomedical informatics from Arizona State University in 2018, Lamp came to UVA Engineering as a Jefferson Fellow. She also joined the Link Lab, where more than 40 faculty and 200 graduate students design cyber-physical systems to transform transportation, cities and health care. There she could collaborate in developing the most advanced cyber-physical systems to give clinicians the best diagnostics for treating disease.
Her advisor, Lu Feng, is an assistant professor with joint appointments in computer science and engineering systems and environment and a member of the Link Lab. Feng connected Lamp to Dr. Sula Mazimba, UVA Medical School associate professor of medicine and cardiovascular disease specialist.
Dr. Mazimba was interested in research that would help better identify heart failure patients at imminent risk of suffering or dying. Lamp’s research focus was particularly relevant. More than 50% of heart failure diagnoses lead to death in less than five years. Access to the most powerful diagnostics in the early stages of the disease would be a real game-changer.
The diagnostics for determining the state of disease in the human heart are complicated. In addition to the measurements of various pressures inside different heart chambers, other metrics characterize the blood flow and amount of oxygen in the blood. All of these metrics in combination form the clearest picture of how well the heart is working. On the other hand, it is difficult to quickly pinpoint what is going on in the heart with so many factors contributing to heart function.
Lamp and Mazimba decided to join forces and create a new scientific method for predicting outcomes based on these metrics, individually and in combination. They hope the research will lead to better overall diagnostics.
“I knew Dr. Mazimba and I would be a good team to determine the best ways to apply science to help these patients,” Lamp said. “Not only is he an incredible clinician, Dr. Mazimba is a professor who understands computer technology and machine learning.”
They are in the process of gathering heart metrics that contribute to heart failure from approximately 400 patients – a total of about 64,000 data points that characterize the overall function of the heart. Lamp will model the data to predict much more specific trajectories of the disease, assign probabilities of bad outcomes and how quickly they may happen.
The analysis forms an elaborate “decision tree” to more accurately assess how far a patient’s heart failure has progressed. Their research will contribute to a patient risk score, with the highest risk a five and the lowest a one. Scores like this can be powerful tools for clinicians needing immediate insight on best treatments.
For example, a risk score of five might indicate the patient has over a 50% chance of dying in the next few months, prioritizing that patient for a heart transplant. A lower risk score might indicate that a patient has time to enter a clinical trial of a novel drug.
“Identifying heart failure patients who are at risk of deteriorating is critically important to the timely allocation of life saving therapeutic interventions," Dr. Mazimba said. "By combining the most advanced computer technology and the most advanced medical practice in our research, we are leading innovation in diagnostics and treatment for the most vulnerable patients.”
Lamp’s dedication to cyber-physical systems research for medicine has not gone unnoticed. She was recently awarded a National Science Foundation Graduate Research Fellowship.
Through the program, the National Science Foundation recognizes graduate students who have demonstrated their potential for significant research achievements and who can also contribute to teaching and innovations in science and engineering.
“Josephine has proven her ability to collaborate with medical technicians in creating functional and secure smart health technologies,” Feng said. “She is a very skilled communicator, and this is a key component in the interdisciplinary research that the Link Lab was created to fuel.”
“For example, the word ‘safety’ can mean something completely different in engineering than it does in medicine, or other disciplines,” Feng said. Engineers frequently refer to safety of cyber-physical systems, meaning their reliability and security. “Learning how to communicate in ways that are meaningful to other collaborators is crucial to success.”
Lamp believes her creativity has been integral to her achievements. While completing her undergraduate studies in bioinformatics, she continued to follow her artistry, minoring in dance. She also worked in the Apple Store fixing computers and in a digital forensics lab performing cybersecurity research.
Lamp shadowed clinicians at the Mayo Clinic as well, seeing first-hand the many areas where formal computer science methods could be a practical tool for improving patient care.“Looking back, my diverse experiences, including artistic ones, have allowed me to be a great engineer and help solve important and interesting problems,” she said.
In a recent presentation, Lamp drew on a life-long love of comics to artfully communicate her work through a superhero analogy that non-scientific audiences could appreciate. She likened the research to “computer science powers for good” used by Spider-Man to defeat the “evils” of heart failure.
Spider Man’s web was made up of the heart metrics for all the human hearts in a large city, with some of those hearts in need of rescue. The “Spidey Senses” were the machine learning and AI-powered modeling tools that help assess the potential for danger. Evil would be defeated by eliminating unknowns to illuminate the best rescue.
Lamp credits UVA Engineering’s emphasis on community as a strong foundation for her research achievements.
“When I visited UVA Engineering, I sensed the collaboration and diversity. The fact that UVA had been recognized as a top engineering school for its percentage of female computer science graduates was also really impressive,” Lamp said. “Ultimately, though, the overriding factor for me in choosing UVA was because it really promoted community.”
That sense of community proved pivotal during the pandemic when faculty, staff and graduate students had to put their heads together for creative solutions to continue their research. The challenge only strengthened the bonds.
“What I treasure is the camaraderie I have here at UVA. I have made friends from all over the world through my research and in the Link Lab.” Lamp said. “I learn so many new things through so many diverse points of view. Even with a world in turmoil, we are able to celebrate our differences and support each other through this challenging time.”
Lamp also serves as the chair of the Computer Science Graduate Student Group, which is a liaison between students and faculty. In that role she organizes events to support student community-building and represents the student body at faculty meetings. “I hope to act as a useful representative and help improve and support the program and students here,” she said.
“Josephine is pursuing incredibly exciting research, and we are exceptionally proud of her accomplishments and her dedication to excellence, as well as her leadership within our community,” said Kevin Skadron, Harry Douglas Forsyth professor and chair of computer science. “The National Science Foundation has recognized her strong contributions to cyber-physical systems research as well as her potential to become a thought leader in safe cyber innovation.”