Finding disease outbreak locations from human mobility data.

Area of Science:

• Epidemiology
• Computational Biology
• Data Science

Background:

• Traditional methods for identifying infectious disease outbreak origins are manual, time-consuming, and prone to errors.
• Existing advanced methods like genomic sequencing or epidemic modeling are not suitable for the initial outbreak phase.
• Digital data, particularly mobile phone data, presents opportunities for automated outbreak source identification.

Purpose of the Study:

• To develop and validate a novel, automated method for determining the origin location of infectious disease outbreaks.
• To leverage geolocated movement data from affected individuals for rapid outbreak source identification.
• To provide a reliable, out-of-the-box solution for early-phase outbreak investigations.

Main Methods:

• The proposed algorithm analyzes geolocated movement trajectories of individuals affected by an outbreak.
• It scans movement data to identify shared locations among affected individuals.
• The outbreak origin is determined as the most dominant shared location identified in the movement data.

Main Results:

• The method accurately identifies the true outbreak location using movement data from a small number of individuals.
• It demonstrates high accuracy across various empirical and synthetic datasets.
• The approach is robust to noise and can handle scenarios with multiple or unknown numbers of outbreak sources.

Conclusions:

• This novel method offers a reliable and accurate approach for identifying outbreak origins in the critical initial phase.
• It significantly improves upon traditional manual methods, enabling faster crisis response.
• The technique is broadly applicable to identifying shared locations in movement data beyond disease outbreaks.