“Foliar diseases are a major and persistent threat to global food security,” explains cereal disease expert, Dr Peter Dracatos. “In barley alone, diseases such as net blotch, leaf rust and scald cause millions of dollars in losses globally each year.”
Screening for disease resistance currently relies on visual scoring – rating disease severity on a 1 to 9 scale – by plant pathologists. This process is time consuming, prone to operator-bias and requires specialist expertise that is in marked decline.
To address this, Dr Dracatos, his PhD student, Matthew Ulrich, and collaborator Dr Ali Zia, are developing the Rust-Expert (RUST-E) pipeline, a high-resolution, image-based disease screening framework.
“RUST-E is designed to emulate and extend the decision-making processes of expert plant pathologists,” PhD student Matthew Ulrich explains.
“The framework embeds biological knowledge directly into AI, ensuring disease traits are defined consistently across experiments, locations and growing conditions.”
“This means more accurate, consistent and scientifically robust decision-making can occur, such as whether or not to spray damaging and costly chemical fungicides.”
Being able to measure small differences in how plants respond to disease will also help scientists to identify previously overlooked sources of partial resistance; essential for developing long-lasting or durable resistance.
“RUST-E will help breeders make better data-driven decisions and speed up genetic improvement as disease pressure increases with climate change.”
