EVALUATING VECTECH IDX™: AI-DRIVEN IDENTIFICATION FOR ENHANCED VECTOR MANAGEMENT

Assessing and advancing cutting-edge technologies that are designed to optimize mosquito surveillance strategies is crucial given the complex challenges presented by our rapidly changing environments. Vectech's Identification-X (IDX) machine offers an innovative way to identify and count adult mosquitoes and train artificial intelligence (AI) software. In collaboration with Vectech, staff at Placer Mosquito and Vector Control District (District) have identified and imaged ∼5,100 adult mosquitoes since 2021. Using the most recent software update (v5.0.4), we aimed to evaluate accuracy improvements by incorporating a diverse range of mosquito genera and species. To achieve this, 100 female specimens of 10 wild-caught mosquito species were imaged. The species included Anopheles freeborni, An. franciscanus, Culex tarsalis, Cx. pipiens, Cx. stigmatosoma, Cx. erythrothorax, Aedes vexans, Ae. melanimon, Ae. aegypti, and Ae. sierrensis. Of the 10 species analyzed, 6 had an identification accuracy of 96-100%. Given the software's current accuracy for Cx. tarsalis, follow-up examinations were conducted to determine at what point consistent specimen degradation impacted the ability of the IDX to correctly identify mosquito samples of this species. Finally, we compared the identification accuracy and speed of individual vector control technicians (VCTs) with the imaging accuracy and speed of the IDX machine to determine operational efficiency of this device within a mosquito surveillance program. Results indicated that the IDX machine is as accurate and efficient as a vector control technician with 22 months of experience but is prone to misidentification of morphologically similar mosquito species when specimens are degraded.