AI4PEP · Lebanon hub

Catching epidemics early, straight from the laboratory

EMERGENT Lebanon uses artificial intelligence to read hospital laboratory results as they are produced, spot unusual rises in priority pathogens, and alert clinicians and public-health teams before an outbreak takes hold.

StrEngthening pandeMic survEillance thRough artificial intelliGENce automation of laboraTory data — for Lebanon.

Weekly confirmed cases · illustrative● Alert · week 16
Expected baselineAlert thresholdFlagged week

Simulated example: bars show weekly cases; one week exceeds the alert threshold and is flagged.

18priority pathogens monitored weekly
7anomaly detectors in one ensemble
1university hospital pilot site
2countries — Lebanon & Canada
Why it matters

Surveillance that keeps working when the health system is under strain

Since 2019, Lebanon's economic crisis, compounded by COVID-19, has weakened public-health infrastructure. Surveillance that depends on manual, self-reported notifications suffers from delays, limited coverage and missing data.

Laboratories already hold the earliest, most objective evidence of infection. EMERGENT connects directly to laboratory information systems so that signals are detected automatically, reviewed by experts, and shared in near real time.

Read about our objectives

Laboratory-first

Data are extracted automatically from hospital laboratory systems using health-data standards (HL7 FHIR, LOINC).

Hybrid AI detection

Statistical and machine-learning detectors work together to flag unusual weekly trends.

Timely alerts

Flags are sent by email and shown on a central dashboard to support rapid decisions.

Experts in the loop

Clinicians and microbiologists review, label and override alerts — the system learns and stays accountable.

How it works

From a test result to an early warning

A simple, transparent pipeline designed with clinical and laboratory stakeholders.

  1. Extract

    Results are pulled automatically from the laboratory information system.

  2. Standardise

    Tests and organisms are mapped to common codes and aggregated into weekly counts.

  3. Detect

    An ensemble of statistical and ML detectors compares each week to what was expected.

  4. Validate

    Experts review the flags, confirm or dismiss them, and tune the alert threshold.

  5. Act

    Confirmed signals reach decision-makers via alerts and a real-time dashboard.

Explore early warning systems & try the demo

Featured publication

AI for Epidemic Preparedness: Innovations and Insights from the Global South

Edited by Christo El Morr, Antoine Saab & Elie Salem-Sokhn · Springer, 2026

An edited volume on applying AI to epidemic preparedness in resource-constrained settings — including a chapter on the design and development of the EMERGENT Lebanon platform.

View on Springer ↗
Recent research

All publications →

News & events

Latest from the project

All news & events →

Partner with EMERGENT Lebanon

Hospitals, laboratories, public-health authorities and researchers interested in scaling AI-driven laboratory surveillance — we would like to hear from you.

Get in touch