USF researchers have launched a social media campaign that invites citizen scientists to upload images of mosquitoes found in the Tampa Bay region to the platform, a smartphone app. The app will automatically identify disease-carrying species such as Aedes aegypti, a known transmitter of Zika, dengue and yellow fever.
This is the next step in a multi-disciplinary effort led by Ryan Carney, assistant professor of integrative biology and Sriram Chellappan, professor of computer science and engineering. The project is funded by a four-year grant from the National Science Foundation. The goal is to build a robust image database that identifies mosquitoes and monitors their habitat by using an integrative approach that requires many helping hands, which Carney calls 鈥渃itizen epidemiology,鈥� where citizen scientists and students participate as data collectors.
鈥淲e need a lot of images for the AI training. Volume is really key, the more data you feed these algorithms, the better they get,鈥� Carney said.
The mosquito images, along with details about their habitat, are essential to train the AI鈥檚 algorithm, which categorizes the anatomical components of the insect and matches them with known characteristics of Aedes aegypti. Over time, this collection of data will better inform mosquito habitat and disease prediction maps with unprecedented detail.
Hillsborough, Pasco, Pinellas and Polk counties are also promoting the photo challenge on their social media channels.
The data will validate a local habitat model of potential risk areas and will later be developed into a Geographic Information System (GIS) dashboard, similar to the (GLOBE) Program鈥檚 website.
Students have been instrumental in building the AI mosquito project.
The social media campaign includes a series of short instructional videos produced by Karlene Rivera, an undergraduate student from the Judy Genshaft Honors College.
鈥淭here鈥檚 definitely a learning curve to producing the videos because of the editing
involved, but it's been a great experience and I鈥檝e learned a lot,鈥� Rivera said.
Rivera鈥檚 videos are helping users learn how to photograph mosquitoes with smartphones,
use the app, and even build mosquito traps. She is also working with communications
teams at and the GLOBE Program to use her TikTok-style videos to invite new citizen scientists
to participate.
On the app side, graduate student has recruited hundreds of iNaturalist superusers, posting updates and mosquito observations to the .
The mosquito habitat model was first developed by Carney鈥檚 former undergraduate student, Connor Mapes. The model was built using mosquito trap data provided by the four mosquito control organizations in Hillsborough, Pasco, Pinellas, and Polk counties during Mapes' thesis work. It helps model the habitat of the Aedes aegypti species and predict risk of mosquito-borne diseases.
鈥淚t鈥檚 really been amazing to be a part of this and grow as a result of this experience,鈥� Mapes said. 鈥淔rom just helping out on a project, to writing my thesis, to having a larger role, has been really awesome.鈥�
Mapes is now working at the and using the datasets collected in the iNaturalist, and Mosquito Habitat Mapper apps to develop a GIS dashboard. The dashboard will allow users to see photos of observed mosquitoes along with habitat details.
Besides working to locate Tampa Bay鈥檚 Aedes aegypti mosquito and its habitat, researchers are also concerned with newer invasive species and future threats in other parts of 91社区 and beyond. Carney worries about the Aedes scapularis, a recent arrival to the Miami-Dade area, and two other species that may potentially invade 91社区 and have already been detected in the Caribbean.
鈥淥ne of the overarching themes of our NSF project is building artificial intelligence and surveillance capacity using citizen science smartphone images, which can be used in the training of the AI,鈥� Carney said.
Research positions for this project are now available for a postdoctoral student, masters student and an undergraduate student. Interested parties can inquire at .