The Division of Health AI at Northwell Health and the Feinstein Institutes develops artificial intelligence and machine learning models to improve patient care and outcomes. We operate at the intersection of artificial intelligence, computational neuroscience, and clinical medicine.
Our work spans multiple healthcare data modalities: electronic health records, continuous physiological monitoring, medical imaging, and neural signal processing. Every algorithm we develop is designed not just to advance the science, but to be deployed in clinical settings where it makes a real difference in patient care.
We combine the rigor of academic research with the urgency of clinical need. Our lab develops AI across five research verticals, each addressing a critical gap in how healthcare leverages data to improve outcomes, from predicting patient deterioration on the ward to decoding neural signals that unlock new therapies.
We don't stop at publishing. We build, validate, and deploy.
Deep learning models trained on 1.5M+ hospitalizations predict deterioration up to 17 hours in advance, outperforming existing systems by 25%. Funded by a $3.2M NIH grant and published in Nature Communications.
AI tools analyzing data from 35,000+ patients to support clinical decisions during the pandemic, published in JAMA and Nature Communications.
Processing 200TB of imaging data from 60 vagus nerves to create the most detailed digital twin in existence. Funded by a $6.7M NIH grant, using 3D nn-U-Net deep learning segmentation.
DeepAR forecasting models predicting nursing demand 12 months ahead, estimated to save $10M annually across 10 hospital units.