Research Mission
Improving neurologically intact survival from cardiac arrest and optimizing the perioperative journey—from prehabilitation through recovery.
Research Areas
The Problem
In-hospital ECPR (extracorporeal CPR) improves survival for refractory cardiac arrest, but transport delays limit its effectiveness. Mobile ECPR could transform outcomes.
Our Approach
Designing and testing prehospital ECPR delivery models through feasibility studies, simulation, and pilot clinical trials comparing prehospital vs in-hospital cannulation.
The Problem
Over 60% of ventricular fibrillation (VF) cases remain refractory to initial defibrillation shocks. Current guidelines follow a one-size-fits-all approach that delays escalation to advanced strategies.
Our Approach
Building foundation models that predict optimal shock strategy (standard dose, vector change, or double sequential external defibrillation) from ECG waveforms during CPR, even with compression artifacts.
The Problem
Anesthesiologists make hundreds of decisions per case. AI assistants could augment clinical reasoning, but rigorous evaluation is needed.
Our Approach
Randomized trials evaluating clinician performance with and without AI assistance on standardized cases, plus digital health platforms for prehabilitation and perioperative optimization.
The Problem
Optimal chest compressions must be centered over the left ventricle. Even a few centimeters of deviation significantly reduces survival, yet current guidelines don't account for patient-specific cardiac anatomy.
Our Approach
Developing real-time feedback systems that use multi-electrode ECG arrays to localize the left ventricle and guide compression positioning.
The Problem
CPR quality and patient physiology are captured in multiple waveforms (ECG, impedance, PPG, capnogram), but this rich data is underutilized for real-time decision support.
Our Approach
Integrating multi-modal signal processing with machine learning to extract actionable insights during resuscitation—from predicting ROSC to optimizing compression parameters.
Methodologies
Key Collaborators
Working with world-class researchers in resuscitation, AI, and biomedical engineering.
Ethics & Open Science
All clinical research undergoes rigorous ethics review. We believe in transparent, reproducible science and commit to:
- •Pre-registration of clinical trials
- •Publication of both positive and negative results
- •Open-source code where possible (respecting patient privacy)
- •Triple-verified citations in all publications