Application of Wings Interferential Patterns (WIPs) and Deep Learning (DL) to classify some Culex. spp (Culicidae) of medical or veterinary importance

In this paper, we test the possibility of using Wing Interference Patterns (WIPs) and deep learning (DL) for the identification of Culex mosquitoes species to evaluate the extent to which a generic method could be developed for surveying Dipteran insects of major importance to human health. Previous applications of WIPs and DL have successfully demonstrated their utility in identifying Anopheles, Aedes, sandflies, and tsetse flies, providing the rationale for extending this approach to Culex. Accurate identification of these mosquitoes is crucial for vector-borne disease control, yet traditional methods remain labor-intensive and are often hindered by cryptic species or damaged samples. To address these challenges, we applied WIPs, generated by thin-film interference on wing membranes, in combination with convolutional neural networks (CNNs) for species classification. Our results achieved over 99% genus-level accuracy and up to 100% species-level accuracy. Nonetheless, challenges with underrepresented species emphasize the need for larger datasets and complementary techniques such as molecular barcoding. This study highlights the potential of WIPs and DL to enhance mosquito identification and contribute to scalable tools for broader surveys of health-relevant Dipteran insects.