At Loyola University Chicago, I built and compared predictive models using 8,737 patient and phantom scans to evaluate markerless tumor-tracking reliability for radiation therapy. My first-author results showed that dual-energy imaging raised AUC to 0.97 for phantom scans and 0.92 for patient scans.
I also quantified image quality and scatter in X-ray images for image-guided radiotherapy, with first-author results in Med Phys. I worked with medical physicists and Varian Medical Systems iLab researchers on a clinical tracking workflow and shared findings with clinical and industry stakeholders.
At Baylor University, I built image-based 3D rat glioma models and ran finite element simulations to study electric field distribution in cancer therapy. Earlier, I partnered with neuropsychologists and surgeons on a clinical study of brain activity during cardiac surgery, collected fNIRS data, and developed MATLAB scripts to analyze the signals.
