Nicole Carpentiere
Pronouns: She/ her/ hers
Research Mentor(s): Karandeep Singh
Co-Presenter:
Research Mentor School/College/Department: Internal Medicine and Learning Health Sciences / Medicine
Presentation Date: April 20
Presentation Type: Poster
Session: Session 2 – 11am – 11:50am
Room: League Ballroom
Authors: Nicole Carpentiere, Norah Saraya, Jonathon McBride, Karandeep Singh
Presenter: 71
Abstract
Objective Studies have shown that even a well-trained human ear is on average only sixty percent effective in diagnosing cardiac problems. Because of this many patients are sent to get an echocardiograph when they might not have needed one. The tests run are expensive and require that the patient be admitted to the hospital. The purpose of this research project is to create a cardiovascular diagnostic tool that would detect cardiac issues at earlier clinical stages and with increased accuracy that would reduce the number of patients sent to get an echocardiograph. Methods Data collection consists of the patients’ echocardiographs along with recordings taken with a digital stethoscope of the patient’s heart at each of the valves. The stethoscope would record a PCG and a one-lead ECG of the heart for a total of thirty seconds at each valve. The goal is to gather data from one thousand patients, and currently, the students have collected over half of the desired total. The participants are eighteen and older and have all received an echocardiograph within ninety days of heart sound collection. All the collected data is then compiled by computer scientists to classify the sounds and ultimately produce the diagnostic program. Results We have collected data on over five hundred patients and we have over two thousand sounds recorded. We will be enrolling more patients in the future to hit the one thousand patients goal. Conclusion Knowing that even the professionally trained ear cannot diagnose heart problems with higher than an average of sixty percent accuracy, it is important to determine if machine learning could improve the accuracy of diagnosis. Furthermore, it would be more affordable and accessible for patients since they would not be sent for unneeded echocardiographs.
Biomedical Sciences, Interdisciplinary
