Can AI-based nematode imaging improve food security?

Plant-parasitic nematodes are a major threat to global food security, causing billions of pounds of crop losses every year. These microscopic, worm-like creatures infect the roots of every major food crop, draining vital resources and reducing harvests. Traditionally, identifying which plants have natural genetic resistance to these pests has been painstakingly slow. But researchers at the University of Cambridge’s Crop Science Centre have developed a method using 3D-printed robots and cutting-edge AI to massively accelerate this work. By automating how they image and analyse nematode-infected plants, the team can now image and analyse around 1,000 plants an hour – a huge increase from the previous limit of just 90 plants a day. The AI models identify nematodes with a high level of accuracy, allowing researchers to track the entire life cycle of the parasite and pinpoint specific plant genes that govern resistance. The ultimate goal of this research is to identify resistance genes that allow for the development of naturally resilient crops and support a future of sustainable farming across the globe. With thanks to Graham CopeKoga at the Cambridge Philosophical Society for producing this video – and to Sebastian Eves-van den Akker and his group including Siyuan Wei, Jie Zhou, Olaf Kranse, Unnati Sonawala, Paul Goodman and George Harpum for their support in documenting their work.