Intelligent Disease Outbreak Risk Assessment Using Generative AI and Collaborative Multi-Agent Systems

Disease-outbreak information is spread across surveillance reports, environmental observations and population statistics,and public-health users need it presented in a form they can inspectand question. This paper presents the design and implementationof an SDG 3 health platform that combines a multi-factor outbreak risk-scoring module, a what-if climate-variation analysis, acollaborative multi-agent chatbot, a locally hosted Mistral 7B Instruct model accessed through LM Studio, a retrieval-groundedknowledge module, and a heuristic hallucination detector for usersubmitted outbreak reports. The risk module combines new cases,rainfall, humidity, temperature, population density and hospitaladmissions into a weighted score from 0 to 100, classifies it as Low,Medium or High, and derives four-week and three-month projections. A React front end communicates with a FastAPI back endthat provides JWT authentication, role-based access, an administrator review workflow with audit events, and report generation. Thelanguage model communicates information only; it does not compute risk scores. The platform does not forecast case counts witha time-series model. A Random Forest classifier for risk-categoryclassification (Low, Medium, High) was evaluated on a balancedtest set of 90 samples and reached 88.89% accuracy and a macroF1-score of 88.89%. The worked examples of the scoring rules arenot accuracy measurements, and learned time-series forecasting,such as LSTM, is identified as future work.Keywords—Disease outbreak risk assessment, generative AI, multiagent systems, Mistral, LM Studio, retrieval-augmented generation,hallucination detection, public health, SDG 3.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-18
DOI
https://doi.org/10.5281/zenodo.24343410
Primary Topic
Data-Driven Disease Surveillance
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Intelligent Disease Outbreak Risk Assessment Using Generative AI and Collaborative Multi-Agent Systems

Ashritha A M, Aishwarya K
1 links
Zenodo (CERN European Organization for Nuclear Research)
Data-Driven Disease Surveillance
article

Intelligent Disease Outbreak Risk Assessment Using Generative AI and Collaborative Multi-Agent Systems

Ashritha A M, Aishwarya K
article en

Abstract

Disease-outbreak information is spread across surveillance reports, environmental observations and population statistics,and public-health users need it presented in a form they can inspectand question. This paper presents the design and implementationof an SDG 3 health platform that combines a multi-factor outbreak risk-scoring module, a what-if climate-variation analysis, acollaborative multi-agent chatbot, a locally hosted Mistral 7B Instruct model accessed through LM Studio, a retrieval-groundedknowledge module, and a heuristic hallucination detector for usersubmitted outbreak reports. The risk module combines new cases,rainfall, humidity, temperature, population density and hospitaladmissions into a weighted score from 0 to 100, classifies it as Low,Medium or High, and derives four-week and three-month projections. A React front end communicates with a FastAPI back endthat provides JWT authentication, role-based access, an administrator review workflow with audit events, and report generation. Thelanguage model communicates information only; it does not compute risk scores. The platform does not forecast case counts witha time-series model. A Random Forest classifier for risk-categoryclassification (Low, Medium, High) was evaluated on a balancedtest set of 90 samples and reached 88.89% accuracy and a macroF1-score of 88.89%. The worked examples of the scoring rules arenot accuracy measurements, and learned time-series forecasting,such as LSTM, is identified as future work.Keywords—Disease outbreak risk assessment, generative AI, multiagent systems, Mistral, LM Studio, retrieval-augmented generation,hallucination detection, public health, SDG 3.

Zenodo (CERN European Organization for Nuclear Research)
East West Institute of Technology (IN)
Good health and well-being
Data-Driven Disease Surveillance
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon