The tsetse fly has wreaked devastation across large swaths of sub-Saharan Africa. About the size of a housefly, the insect transmits a parasite that can be lethal to both humans (sleeping sickness) and animals (nagana). A Columbia team is therefore pioneering a machine-learning based imaging and sorting solution that aims to drastically reduce Africa's tsetse population.
Columbia research team uses data science to eliminate the tsetse fly in sub-Saharan Africa.
Products and exhibitors from the field of AI and ML
Machine Learning is an exciting field that also opens up new possibilities for medical technology. You can find exhibitors and products related to Information and Communication Technology in the catalogue of MEDICA 2018:
The research is supported by a $100,000 Seeds Fund Grant from Columbia's Data Science Institute. The solution, which allows for the sorting of male and female tsetse flies, was discovered by Zelda Moran, now a researcher at the Earth Institute. While working as an intern at the International Atomic Energy Agency (IAEA), Moran first used near infrared light to peer inside the pupae of tsetse flies. She placed the pupae under a microscope and used the near infrared light to photograph them, allowing her to discern if the developing flies were male or female. She designed this novel sorting technique to support the Sterile Insect Technique, which the IAEA has used to eradicate tsetse populations in Zanzibar and other countries. The sterilizing technique uses irradiation to render large numbers of male flies infertile. The flies are then released into breeding grounds, where they mate with female flies. Since the females usually mate only once in a lifetime, the unfertilized mating will drastically reduce the tsetse population and help eliminate the spread of the disease.
As it is, though, fly-production labs use a manual and time-consuming technique to sort the flies by sex. The sorting solution pioneered by the Columbia team, once refined, will enable the labs to sort the flies faster and more efficiently and on a larger scale. This enhancement will allow the labs to ship a higher number of sterile flies, in earlier stages of their life cycles, out to the sub-Saharan breeding grounds.
After Moran finished her internship at IAEA and later moved to New York, she teamed up with the Marka lab at Columbia. In that lab, Szabolcs Marka, an astrophysicist with an expertise in biophysics, and Zsuzsa Marka, an experimental physicist, had previously collaborated on insect behavior research, including how to mitigate malaria mosquitoes and fruit flies. The Markas' expertise in optics, hardware instrumentation and electronics has also been used successfully in biophysics and astrophysics applications. Most prominently, they are members of the LIGO Scientific Collaboration, a consortium of scientists who worked on the instrument that in 2015 detected gravitational waves. John Wright, the fourth member of this research team, is an electrical engineer who does high-dimensional data analysis and develops algorithms to solve imaging problems-in this case, the pupae images. Wright and the Markas are also affiliates of the Data Science Institute. Together, the team aims to work in the lab (find out more about Laboratory Technology at MEDICA 2018 here) to develop a sorting-machine prototype that can process tens of thousands of pupae images a day.
Right now, the four are focused on amassing large numbers of pupae images that will be used to train a machine-learning algorithm. At the same time, they are also designing a robot that can sort the pupae based on the result of the algorithm. Once the robot can process a large number of images with high accuracy, they will send it to the International Atomic Energy Agency, which will use it to sort the male flies, irradiate them and release them in Africa.
"We'll use data science to create a machine-learning-based automated system (find out more about Information and Communication Technology at MEDICA 2018 here) that can quickly determine from a tsetse pupae if it's male or female," says Zsuzsa Marka. "And that system will be the basis for a robotic sorting machine that we hope will eventually be used in sub-Saharan Africa to mitigate the scourge of the tsetse fly."
MEDICA-tradefair.com; Source: Data Science Institute at Columbia