Infection Inspection
A new way to analyze disease transmission simulations.
Public health officials agree that the end of the COVID-19 pandemic is in sight. However, the world has adjusted to a state of constant vigilance for not only COVID flare-ups, but other potential deadly pathogens as well. Many still feel a lingering sense of danger in the wake of the pandemic’s massive societal ramifications, and pose the question: how do we prepare ourselves for inevitable future outbreaks?
Enter the epidemiologist. Often called “disease detectives”, these public-health workers dedicate their lives to searching for the cause of disease, identifying people who are at risk, and investigating strategies to control or stop its spread.
Epidemiologists rely heavily on data visualization to simulate disease spread over dynamic contact networks, a technique more commonly known as infection mapping. However, the current technology used to interpret these simulations is confined to conventional statistical analysis.
This methodology is restrictive when applied to more complex transmission patterns and was especially limiting during the pandemic. As a result, development to solve the issue has been rapidly materializing—culminating in a novel technique that visualizes infection maps using the concept of representative trees.
Representative trees cluster local disease outbreak patterns based on their similarity in disease spread and allow for an aggregated overview, while still preserving epidemic structures for large infection maps.
This unique approach to infection mapping facilitates richer and effective interactions. It also allows epidemiologists to better understand the results of simulated models, and identify more issues within them.
The concept of representative trees was constructed from the ground up by a group of interdisciplinary research professors in the United Kingdom, who recently published a scholarly article documenting their initial research and testing process.
Notable aspects of this process included testing use cases for representative trees, performing a qualitative expert evaluation, conducting semi-structured interviews, and hosting a think-aloud seminar whereby participants in the field were asked to talk through their analytical thinking and reasoning in relation to their actions.
After completing these four research stages, the creators compiled a list of future plans for their research. Specifically, they mentioned how the ability to concurrently compare data at multiple levels of aggregation would be a key visualization strength to implement.
They also acknowledge that, while the tool is built for simulated data, it should be possible to eventually adapt it to real infection maps from actual contact tracing data.
On a broader scale, the development and testing of representative trees as a novel method for infection mapping signifies the importance of delivering effective and thoughtfully designed visual analytics approaches to support complex situations.
References
- Sondag, M., et al. “Visual Analytics of Contact Tracing Policy Simulations During an Emergency Response.” Computer Graphics Forum, 41(3):29–41, 2022, https://onlinelibrary.wiley.com/doi/10.1111/cgf.14520.
