Endometriosis affects 5-10% of reproductive-aged women but faces diagnostic delays of 6-11 years, contributing to prolonged suffering and healthcare inequities. Current diagnostic approaches rely on single data sources and fail to capture the complex, multifaceted nature of this disease. This project develops a multimodal machine learning framework that fuses three complementary data streams: structured longitudinal electronic health records, patient-generated symptom reports from a menstrual-tracking app, and clinical medical imaging. By integrating these heterogeneous modalities into a unified predictive model, we can leverage the strengths of each data source to enable earlier and more accurate risk prediction. Critically, our approach incorporates fairness auditing and debiasing techniques to address documented disparities in endometriosis care across race, age, and socioeconomic groups. We hope to transform patient outcomes by shortening diagnostic delays and enabling timely, equitable interventions for all patients.

Through this SuperUROP, I aim to leverage my background in computer science to develop innovative solutions for a critical healthcare challenge. I’m particularly excited to apply machine learning to improve early detection of endometriosis while addressing fairness and equity in clinical AI systems. I’m drawn to problems that bridge multiple disciplines and that close the gap between algorithmic innovation and real-world outcomes with potential clinical applications.