Skip to content
Division of Health AIDivision of Health AI
AboutTeamResearchPublicationsInternship
AboutTeamResearchPublicationsInternship
Division of Health AIDivision of Health AI

Clinical AI built with the data and clinicians of Northwell Health.

Explore

  • About
  • Team
  • Research
  • Publications
  • Internship

Affiliations

  • Feinstein Institutes↗ (opens in new tab)
  • Northwell Health↗ (opens in new tab)
  • Zucker School of MedicineHofstra Northwell

Located at

  • Institute of Health System Science
  • Institute of Bioelectronic Medicine
  • Manhasset, New York

© 2026 Division of Health AI, Northwell Health. All rights reserved.

The Division of Health AI

About

Led by

Theodoros Zanos, PhD→

Head of Division
Feinstein Institutes
Northwell Health

The Division of Health AI builds machine learning models that run inside Northwell Health, from early-warning systems at the hospital bedside to decoders for vagus nerve signals. The work combines computational neuroscience, clinical medicine, and software engineering.

Machine learning built inside the hospital system it runs in.

The work is done with the clinicians who will use it, on Northwell’s own records. The deterioration model already runs in silent mode inside the hospitals’ Epic system.

Institution
Feinstein Institutes for Medical Research
Health system
Northwell Health
Location
Manhasset, New York
Focus
Clinical, bioelectronic, and operational AI

The lab over time

One group photo a year, from 2017 to today.

Lab team in 2017, five members in the office
Lab team in 2018, outdoor group photo with team interns
Lab team in 2019, patio lunch
Lab team in 2021, rooftop with the Manhattan skyline
Lab team in 2023, rooftop at dusk with the NYC skyline
Lab team in 2024, rooftop at sunset with the NYC skyline
Lab team in 2025, garden group photo with inset headshots along the top

Lab team in 2017, five members in the office.

Meet the current team→

What the lab has built

Four programs built inside Northwell, each with published results.

Clinical AI

Patient deterioration prediction

A wearable-based deep learning model, published in Nature Communications, flags in-hospital deterioration up to 17 hours before onset from just nine physiological signals. Supported by a $3.1M NIH grant, it anticipates rapid response calls, ICU transfers, and other adverse outcomes.

Pandemic Response

COVID-19 clinical decision support

Machine learning tools developed during the pandemic to support clinical decisions for COVID-19 patients, drawing on tens of thousands of cases across multiple cohorts. Published in JAMA and Nature Communications.

Bioelectronic Medicine

Vagus nerve digital twin

High-resolution 3D reconstructions of 60 human vagus nerves segmented from hundreds of terabytes of imaging data using nn-U-Net. Part of a $6.7M NIH SPARC award.

DEMANDATTRITION

Operational AI

Workforce forecasting

DeepAR probabilistic forecasting of nursing demand at 12-month horizons across Northwell's hospital units, supporting preemptive staffing decisions.

See all research→

More

Explore the researchRead the papers→