AI-based epidemic and pandemic early warning systems: A systematic scoping review

Area of Science:

• Public Health
• Epidemiology
• Computer Science

Background:

• Timely detection of disease outbreaks is critical for effective public health response.
• Artificial Intelligence (AI) offers potential for identifying patterns indicative of epidemics and pandemics.
• Early Warning Systems (EWS) are essential for monitoring and responding to public health threats.

Purpose of the Study:

• To evaluate the effectiveness of AI in epidemic and pandemic early warning systems.
• To identify challenges and propose strategies for improving AI-based EWS.
• To assess the predictive capabilities of AI systems in public health surveillance.

Main Methods:

• A systematic scoping review methodology was employed.
• Studies published within the last five years focusing on AI/machine learning in EWS were included.
• Thematic analysis was conducted on 33 selected articles after screening 1087.

Main Results:

• AI-based EWS have demonstrated effectiveness across diverse applications and algorithms.
• Identified challenges include data quality, model explainability, bias, and data characteristics (volume, velocity, variety, availability, granularity).
• Strategies for mitigating AI bias and enhancing system adaptability were explored.

Conclusions:

• AI shows significant promise in improving the speed and accuracy of epidemic detection.
• Addressing data quality, bias, and model transparency is vital for enhancing the reliability and generalizability of AI-EWS.
• Future development requires continuous monitoring, improvement, and integration of socio-environmental data.