United States
Building weakly supervised models that detect and label critical anatomical structures in microscopic neurosurgical videos Developing a multimodal retrieval-augmented LLM to support neurosurgeons in complex scoliosis decision-making
Conducted research at the Rauschecker-Sugrue Lab, Center for Intelligent Imaging (ci²), Department of Radiology & Biomedical Imaging, University of California, San Francisco (UCSF), under the supervision of Dr. Andreas Rauschecker Focused on 3D meningioma segmentation from T1-post contrast MRIs, achieving a 0.93 median Dice score on manually labeled data from Duke and UCSF, and on quantifying model uncertainty for clinical deployment to enhance reliability
- Guided and mentored 26 students within the Cordées de la Réussite government program. - Organized and led groups of three students, assessing their oral communication skills to track progress.