This summer, I conducted research under Dr. Stuart Corr at the Houston Methodist Innovation Engineering Core, with additional mentorship from Dr. Omar Ahmed and Dr. Salman Khan. My project focused on developing a CT-based machine learning system to predict whether a patient undergoing endoscopic sinus surgery would require a septoplasty to create sufficient surgical access. Using just 57 preoperative CT scans, we quantified septal deviation and nasal airway geometry and developed an interpretable model that achieved a promising AUC of 0.78, supporting further validation in a larger patient cohort.

This work could provide surgeons with a more objective way to plan septoplasty before surgery, reducing unexpected intraoperative decisions, procedure time, anesthesia exposure, and unnecessary recovery for patients who may not need the additional procedure.