AI pinpoints Origins of medicinal root to combat fraud
Summary and headline written by AI from the source article. How we work
Researchers in China used machine learning to determine the geographic origin of Gastrodia elata, a root used in traditional medicine and as a functional food, with over 92 percent accuracy.
A team led by Dan Zhao collected 270 samples of the root, known as tianma, from across China and analysed their chemical composition. They combined measurements of elements from the soil with the levels of active compounds within the root to create a unique “fingerprint” for each region.
The study addressed a growing problem of fraudulent labeling in the market for this increasingly popular ingredient. Regions with established reputations for high-quality Gastrodia elata, like Zhaotong and Dafang, command higher prices, leading to instances of mislabeled and inferior products being sold as premium goods. The researchers focused on the Hongtianma cultivar, which receives less attention than other varieties despite being the most commonly cultivated.
After testing nine different machine learning algorithms, a support vector machine proved most effective at identifying the origin of the roots. The analysis showed that elemental composition was a stronger indicator of origin than the levels of medicinal compounds. The team also found connections between the root’s chemical makeup and factors like rainfall and temperature, suggesting that environmental conditions significantly influence its composition. They plan to expand the research with more data collected over multiple years.
