An Adaptive Multimodal Predictive Intelligence System Integrating Genomic Data, Mobility Patterns, and Public Health Reports for Real-Time Epidemic Outbreak Management

• Ada John

  ladoke Akintola university of technology

  Author

Contemporary epidemic outbreak management remains constrained by fragmented surveillance systems that rely on single-modality data sources, generating critical delays in outbreak detection and response mobilization. This study addresses the fundamental gap in integrated predictive intelligence by developing and validating an Adaptive Multimodal Predictive Intelligence System (AMPIS) that fuses genomic surveillance data, population mobility patterns, and public health surveillance reports into a unified real-time forecasting architecture. Using a design-based research methodology combining retrospective data analysis with prospective simulation across 18 months of multi-country epidemic surveillance data, the study evaluated AMPIS against established unimodal and ensemble baselines. The proposed system achieved 89.4% predictive accuracy (F1-score: 0.891) for outbreak surge detection at a 14-day forecast horizon, outperforming genomic-only models (72.3%), mobility-based models (68.7%), and traditional ensemble approaches (81.2%). Feature importance analysis revealed genomic mutation velocity (0.34), mobility reduction indices (0.28), and syndromic report acceleration (0.22) as the strongest predictors. The findings demonstrate that adaptive multimodal fusion substantially improves both prediction accuracy and lead time compared to existing surveillance paradigms, providing a replicable framework for resource-constrained public health agencies seeking to operationalize digital epidemiology infrastructure.

• pdf

Copyright (c) 2026 Ada John (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.