Projects

Senior AI Architect

Engineered a 3D reconstruction pipeline for cone-beam computed tomography, enabling clinical imaging on refurbished hardware at a fraction of new-equipment cost. Developing machine learning models for orthopaedic surgical planning and ligament modeling, in prototype use by surgeons and progressing toward FDA clearance. Architected the company's quality management system and FDA regulatory submission pipeline.

Lead developer · Gary J. Miller Orthopaedic Biomechanics Lab

The first software to autonomously measure joint kinematics with clinical accuracy from single-plane fluoroscopic images. Used by roughly ten research laboratories across the US, Japan, Canada, and Belgium. My contributions span the full pipeline: CNN-based implant segmentation, GPU-parallelized shape-based pose estimation, and autonomous multi-tier optimization for pose refinement.

Ph.D. research · advised by Dr. Scott Banks · University of Florida

Dissertation: "Autonomous Methods of Measuring Native and Implanted Joint Kinematics from Single-Plane Fluoroscopic Images." Built the estimation pipeline behind JTML: convolutional neural networks for implant segmentation, normalized-Fourier-descriptor pose initialization (GPU-parallelized with a 20× speedup), and contour-based DIRECT optimization validated against three independent bi-plane kinematics datasets with no statistically significant differences.