Density-equalizing Euclidean minimum spanning trees for the detection of all disease cluster shapes

Area of Science:

• Epidemiology
• Geographic Information Systems (GIS)
• Graph Theory
• Biostatistics

Background:

• Current disease cluster detection methods struggle with identifying clusters of diverse shapes and sizes.
• Existing techniques may inaccurately represent the true geographic extent of disease outbreaks, particularly for irregular clusters.
• Overestimation of cluster boundaries by current methods can lead to inefficient resource allocation and public health interventions.

Purpose of the Study:

• To introduce a novel graph-theoretical method for detecting arbitrarily shaped disease clusters.
• To overcome the limitations of existing methods in identifying irregular cluster shapes and sizes.
• To improve the accuracy and sensitivity of disease cluster detection, especially for non-circular outbreaks.

Main Methods:

• Utilized a graph-theoretical approach based on the Euclidean minimum spanning tree.
• Employed cartogram transformation of case locations to represent spatial data.
• Applied the method to historical disease outbreak data, including West Nile virus and inhalational anthrax.

Main Results:

• The proposed method effectively detects arbitrarily shaped disease clusters.
• Performance was comparable to existing methods for approximately circular clusters.
• Demonstrated significant improvement in detection accuracy for non-circular and irregular disease clusters.
• Accurate identification of cluster extent, avoiding overestimation.

Conclusions:

• The graph-theoretical method offers a superior approach for disease cluster detection, particularly for complex outbreak shapes.
• This method enhances the precision of spatial epidemiology by accurately delineating disease extents.
• The findings suggest a potential advancement in public health surveillance and response strategies for infectious diseases.