Researchers reporting in Analytical Chemistry have developed a portable electrochemiluminescence microarray sensor paired with machine learning analysis to quantify zearalenone, a mycotoxin produced by Fusarium fungi that contaminates grains and animal feed.
The device is designed for field-deployable detection at sensitivities relevant to regulatory thresholds. The authors report that the machine-learning component improves quantitative accuracy over conventional readout methods on the same sensor platform.
Commercial mycotoxin testing has traditionally relied on benchtop instruments and laboratory infrastructure. Portable formats, if they hold up under real-world conditions, could shift screening earlier in the supply chain and bring detection closer to production sites and food manufacturers.