AI-Driven Surveillance of West Nile Virus in Maryland: Integrating Vector Identification with Environmental and Epidemiological Insights

This study presents an integrated operational surveillance effort conducted in Anne Arundel County, Maryland, during the 2023 and 2024 mosquito seasons, revealing substantial variation in Culex pipiens s.l. abundance and West Nile virus (WNV) infection risk across time and space. In 2024, mosquito abundance and WNV-positive pools increased more than four- and five-fold, respectively, compared to 2023, underscoring the dynamic nature of local vector populations and arboviral transmission. Observationally, temperature was moderately predictive of mosquito abundance with a 1–2 week lag. Yet, correlational analyses only revealed a relationship between mosquito abundance and precipitation with a 3-week lag in 2023. Notably, ᐩthe temporal overlap between peak mosquito abundance and peak WNV infection was more synchronized in 2024, potentially heightening human transmission risk. These trends informed targeted vector control operations by the Maryland Department of Agriculture, demonstrating the importance of high-resolution, temporally responsive surveillance systems. This work also highlights the operational value of the IDX vector identification platform as part of a scalable entomological surveillance workflow. The mosquito imaging and identification capabilities of IDX supported sample triage, cold chain preservation, and molecular testing without delay. Its integration into the surveillance pipeline enhanced data integrity by enabling efficient specimen logging and species-level confirmation across thousands of samples. The combination of field collection, automated identification, and WNV testing illustrates a deployable model for responsive vector surveillance. This approach supports swift decision-making and demonstrates a modernized framework for mosquito control programs aiming to align operational capacity with climate-driven risk dynamics and advanced technologies.