Healthcare systems and operations are complex and often not running efficiently. To improve them, we partner with clinicians and healthcare managers to identify bottlenecks in the systems and use optimization, data analytics, and operations research tools to solve the identified problems. Our projects expand from the Emergency Department to various hospital departments (e.g., General Medicine, Surgery) to primary care and outpatient clinics.
Identifying the best treatment plan for patients can be complex and high-dimensional by nature and this complexity is amplified by the increasing integration of technology in medicine. We these projects, we consider the problem of finding good treatment plans for cancer patients who undergo radiation therapy using optimization models and algorithms. A successful treatment plan delivers an appropriate amount of dosage to cancerous cells while sparing the healthy surrounding tissue as much as possible.
Inverse Optimization and Inference models
In many complicated systems the outcome depends on a complex set of variables and is subject to a set of constraints. While the final outcomes can be observed (e.g., the decisions made by an expert or a clinician), the exact criteria and properties that led to that outcome are unknown. Inverse optimization is a methodology that is used to infer the underlying optimization problem that decision-makers implicitly consider, often based on past experience or expert knowledge.