Application of Deep Learning to Community-Science-Based Mosquito Monitoring and Detection of Novel Species

Abstract

Mosquito-borne diseases account for human morbidity and mortality worldwide, caused by the parasites (e.g., malaria) or viruses (e.g., dengue, Zika) transmitted through bites of infected female mosquitoes. Globally, billions of people are at risk of infection, imposing significant economic and public health burdens. As such, efficient methods to monitor mosquito populations and prevent the spread of these diseases are at a premium. One proposed technique is to apply acoustic monitoring to the challenge of identifying wingbeats of individual mosquitoes. Although researchers have successfully used wingbeats to survey mosquito populations, implementation of these techniques in areas most affected by mosquito-borne diseases remains challenging. Here, methods utilizing easily accessible equipment and encouraging community-scientist participation are more likely to provide sufficient monitoring. We present a practical, community-science-based method of monitoring mosquito populations using smartphones. We applied deep-learning algorithms (TensorFlow Inception v3) to spectrogram images generated from smartphone recordings associated with six mosquito species to develop a multiclass mosquito identification system, and flag potential invasive vectors not present in our sound reference library. Though TensorFlow did not flag potential invasive species with high accuracy, it was able to identify species present in the reference library at an 85% correct identification rate, an identification rate markedly higher than similar studies employing expensive recording devices. Given that we used smartphone recordings with limited sample sizes, these results are promising. With further optimization, we propose this novel technique as a way to accurately and efficiently monitor mosquito populations in areas where doing so is most critical.

Wing Interferential Patterns (WIPs) and machine learning, a step toward automatized tsetse (Glossina spp.) identification

“…The need for community participation in blood-sucking invasive species identification, for example, Aedes (Ochloretatus) albopictus, Skuse 1895 has pushed deep learning methodology in the entomological survey field. An increasing number of studies are published, focusing solely on this invasive species with identification challenges on imago [33][34][35][36] or larval stage 37 . In addition, the design of traps with embedded systems for counting trapped insects opens up possibilities for real-time surveillance of insect density, a crucial parameter in the survey of insect vectors of medical or veterinary interest 38 .…”

Applications and advances in acoustic monitoring for infectious disease epidemiology

“…In relation to human disease, some research has been conducted into using acoustic sensing in community-science-based mosquito surveillance [10]. Mobile phone devices can be adapted as acoustic sensors to track human-mosquito encounters and, notwithstanding some limitations in implementation, collect occurrence data without typical sampling biases to inform vector-borne disease control programmes [11][12][13]. In a laboratory context there are also examples of acoustic data used to understand mating behaviour of mosquito vectors [14].…”

Measuring factors affecting honey bee (Hymenoptera: Apidae) attraction to soybeans using bioacoustics monitoring

“…Bioacoustics is a branch of science that focuses on sound production by living organisms, and bioacoustics monitoring is a relatively new method for the detection and identification of species through audio recording and analysis. Deep learning approaches to bioacoustics monitoring are currently being developed for the detection and identification of mosquitoes that vector malaria ( Vasconcelos et al 2019 , 2020 , Kiskin et al 2020a , 2020b , Hassall et al 2021 , Khalighifar et al 2021 , Kim et al 2021 ) and invasive fruit fly pests ( Kalfas et al 2022 ). Bioacoustics monitoring has also been combined with machine learning to automate the analysis of birdcall recordings and increase bird monitoring efficiency through the use of the deep neural network BirdNET ( Kahl et al 2021 , Toenies and Rich 2021 , Wood et al 2021 ).…”