Student Research Reports
Open social media sharing options menu
Aedes, Anopheles, Culex: Utilizing Convolutional Neural Networks to Identify the Environment Where Each Mosquito Genera Thrives
Country:United States of America
Student(s):Fathia Bakare, Barok Gebre, Ajani Grant
Grade Level:Secondary School (grades 9-12, ages 14-18)
GLOBE Teacher:Cassie Soeffing
Contributors:Dr. Rusty Low, SME, IGES. Dr. Di Yang, SME, UWyo. AJ Caesar, peer mentor. Peder Nelson, SME, OSU. Andrew Clark, SME, IGES. Dr. Erika Podest, SME, NASA JPL.
Report Type(s):International Virtual Science Symposium Report, Mission Mosquito Report
Protocols:Land Cover Classification, Earth As a System, Mosquitoes
Presentation Video:
View Video
Presentation Poster:
View Document
Optional Badges:I am a Problem Solver, I am a Collaborator, I am a STEM Storyteller
Language(s):English
Date Submitted:01/17/2024
Although mosquitoes may be vital to the animal food chain, they threaten the quality of life and survival of humans around the globe, especially in developing nations. Mosquitoes transmit deadly vector-borne diseases to humans and livestock, killing an estimated 700,000 humans per year according to US Centers for Disease Control and Prevention. One way to mitigate the impact of mosquitoes on human communities is to first identify which types of mosquitos threaten types of communities, as each of the three major mosquito genera (Aedes, Anopheles, and Culex) transmit different diseases - Aedes transmit Dengue and Zika, Anopheles transmit Malaria, and Culex transmit West Nile Virus - so different communities require different treatments for vector-borne diseases depending on the mosquitos that threaten their area. This research utilizes Mosquito Habitat Mapper (MHM) observations made by citizen scientists in the Americas, Africa, and East Asia using the NASA-funded GLOBE Observer mobile application to determine which genera of mosquito is most prevalent in rural, suburban, and urban communities in each region. Each of the 13,000 observations analyzed in this study includes both a photograph of the habitat in which the mosquito larvae were found and a photograph of the mosquito larvae that depicts the mosquito genera that it belongs to. The AI model Convolutional Neural Networks (CNN) was used to classify the habitat as either Rural, Suburban, or Urban and the corresponding mosquito larvae as Aedes, Anopheles, or Culex with over 65% accuracy each. Once habitat and mosquito larvae photographs were classified for all 13,000 observations, a chi-squared test was performed on the data which found that there was a statistically significant relationship between environment and mosquito genera (p < 0.001). By then comparing the expected and observed counts, researchers were able to identify which genera of mosquitoes are both overabundant and under-abundant in each type of environment.