AI-Driven Disease Vector Identification Platform Addresses Key Capability Gap

D. Pecor
Walter Reed Biosystematics Unit, Maryland, United States

Keywords: Convolutional neural networks, vector identification

Effective vector borne disease risk management relies on robust vector surveillance informing FHP decisions. However, vector identification is labor-intensive, and expertise is difficult to build and maintain within the force. To address this critical capability gap, the Walter Reed Biosystematics Unit partnered with VecTech Inc. to develop the IDX Tower platform for identifying vectors using high resolution imagery analyzed using convolutional neural networks to rapidly identify specimens and flag them for pathogen screening more accurately and efficiently. This capability is now being used across a global network of laboratories to identify vector threats to human and animal health. The Warfighter may encounter thousands of different varieties of arthropods while deployed around the globe. Only a very few pose a significant threat to mission success. However, to efficiently identify these threats in time to mitigate potential impacts to readiness, the Warfighter is required to maintain a complex skillset to identify each threat, wherever they are deployed. We aim to demonstrate how this technology is currently being used and discuss future applications as well as a live demonstration of the system.