A prospective program combining wearable ECG and physiologic sensing, autonomic testing, and cardiac substructure dosimetry to characterize cardiovascular and autonomic response during thoracic radiotherapy. The long-term goal is early identification of treatment-related cardiovascular dysfunction using low-burden measurements that can extend beyond the clinic.

A collaboration leveraging the All of Us Research Program to study longitudinal physiologic changes surrounding cancer treatment using wearable and electronic health record data. The project focuses on scalable digital biomarkers — including activity-normalized heart rate, tachycardia burden, and related physiologic measures — that may reveal early treatment-related cardiovascular stress at population scale.

Using routinely acquired cone-beam CT during lung cancer chemoradiotherapy to characterize how tumors evolve during treatment, identify distinct response phenotypes, and determine whether early regression patterns predict recurrence, survival, and other clinically meaningful outcomes.

Developing dynamic prediction and decision-making frameworks that combine evolving biological and physiological response with radiation exposure to cardiac substructures. The goal is to identify when cardiovascular risk meaningfully changes during treatment and when additional surveillance, cardioprotective intervention, or treatment adaptation should be considered.

Developing optimization frameworks that incorporate biologic, imaging-based, and machine-learning-derived predictions directly into treatment planning. The emphasis is on translating predictive models into actionable plan changes while preserving dosimetric quality, clinician control, and robustness to model uncertainty.

Developing computational strategies for tailoring radiotherapy to treatment-induced immunologic response following preoperative immunotherapy. The project asks whether biologic response to systemic therapy can be used to redesign subsequent radiation treatment while preserving disease control and reducing normal-tissue exposure.

Developing a mathematically principled framework for monitoring lymphocyte dynamics during cancer treatment. The project combines compartmental modeling, observability analysis, state estimation, and optimal experimental design to determine what can be reliably inferred from blood measurements and when those measurements should be collected.

An optimization-based platform for coordinating treatment start dates, machine assignments, urgency, treatment complexity, and limited clinical capacity. The program is designed to standardize scheduling decisions while improving access, utilization, and operational efficiency across radiation oncology.

A wearable-enabled mobile platform for longitudinal assessment of cardiovascular and physiological function during and after cancer therapy. The platform is designed to reduce monitoring burden, integrate patient-generated data with clinical research, and create a pathway toward remote, patient-centered digital health interventions.