USF researchers have been awarded a $900,000, four-year grant from the National Science Foundation to use artificial intelligence to fight mosquito-borne diseases. Ryan Carney, assistant professor of integrative biology, and Sriram Chellappan, professor of computer science and engineering, will lead a multi-disciplinary effort to collect data through smartphone apps such as NASA's GLOBE Observer Mosquito Habitat Mapper. App users will be able to upload images of mosquitoes from anywhere in the world and automatically identify their species and potential to transmit viruses, such as West Nile, dengue, and Zika. Mosquito-borne diseases kill almost 3 million people every year.
"It's really a grassroots approach to leverage the computers that we all carry around in our pockets, smartphones, to fight the deadliest animal on the planet," Carney said.
Real-time surveillance is critical for effective mosquito control, currently the only way to combat mosquito-borne diseases. This work scales predictive models out to communities, while also building a global database that tracks the lifecycle of mosquitoes. It's an integrative approach that requires many helping hands, which Carney calls "citizen epidemiology," where citizen scientists and students participate as data collectors.
"These robust models are better tools for citizens to contribute data and reach county epidemiologists to have better disease forecasts," Chellappan said.
As images are uploaded, Chellappan's algorithms categorize anatomical components of the insect and match them with known characteristics of each species. Over time, this collection of data will better inform mosquito habitat and disease prediction maps with unprecedented detail. Artificial intelligence operators will monitor the algorithm periodically to retrain the model and match human expertise.
The collaboration includes public education coordinators and mosquito control districts who will use this technology in K-12 classrooms so that students can learn mosquito biology and contribute to data collection.
The USF scientists have already categorized over 30,000 images in their preliminary work with the Hillsborough County Mosquito Management Program, and plan to add 200,000 more images to their AI program. They hope to turn this collaboration and sharing of data into a global surveillance system that will locate transmission hot spots earlier and improve the control of mosquito-borne diseases.
"What we're trying to do is build this infrastructure so that we can tackle ongoing diseases, but also be better prepared for 'the next Zika'," Carney said.
Carney and Chellappan are currently looking to expand their teams with graduate students and a postdoctoral researcher. Learn more about their project by visiting CarneyLab.org and csee.usf.edu/~sriramc.
