157 - An AI imaging attachment for remote automated assessment of vector abundance and species diversity in mosquito traps

Vector surveillance is critical to understanding mosquito abundance, species diversity, and distribution. When fast and reliable surveillance data is available, mosquito control programs can effectively optimize interventions and assess impact. Unfortunately, conventional surveillance programs require significant labor, taxonomic expertise, and logistics to deploy traps on a daily basis, retrieve specimens, and sort, identify, and record specimen data. Recent advancements of AI algorithms known as convolutional neural networks (CNNs) have enabled image recognition tools, such as Vectech’s IDX, to augment species identification in the lab. This presentation reports on the translation of such AI imaging systems to a tool called Scout designed to attach to mosquito traps. Scout immobilizes and images mosquitoes as they enter a trap, extracts abundance and species data from the images, and transmits the information to a web dashboard in real time. This presentation will describe the attachment design, initial lab test results, and field testing planned in summer 2024.