This was a project I worked on at Johns Hopkins with my buddies Martin Kang (Mount Sinai), Pat Bovonratwet (Yale), and Kyle Xiong (Carnegie Mellon), as well as Professor Marc Donohue (Johns Hopkins). We developed a PK-PD model for using Chemotherapy against tumors. We then seeked to optimize the dosing regimen under a simple objective function to minimize both the mass of the tumor and any collateral damage. It was a fun research project incorporating various compartment models as well as some optimal control theory.
Now that I’m a PhD student at Duke, I wonder if any Reinforcement Learning or Bayesian approach would work for this project. More to come soon!