Determining the Geographical Origin of Gastroda elata f. glauca

The food and pharmaceutical industries rely on transparency in order to inspire confidence in their products. As part of the production processes, industry manufacturers examine and authenticate every product, ensuring its high quality before being shipped out to market. A recent study in Food Control explored this issue recently by looking at Gastrodia elata f. glauca (GEFG), an important medicinal and food plant in China (1).

This study was led by Yuanzhong Wang and Honggao Liu from the Yunnan Agricultural University Medicinal Plants Research Institute, Yunnan Academy of Agricultural Sciences, and Zhaotong University. In the study, the research team used Fourier transform infrared (FT-IR) spectroscopy with deep learning to enhance GEFG traceability (1).

Orchid flower. | Image Credit: © phanthit malisuwan - stock.adobe.com

GEFG is an herb in the Orchidaceae family (2). It is known for its medicinal and nutritional benefits. GEFG is mainly found in the Changbai Mountain region, which is in eastern China (3). Previous studies (2,3) have explored the utility of GEFG to help alleviate specific health symptoms. This study concentrated on analyzing the dry matter content (DMC) of GEFG in the Yunnan province as well as other provinces, examining environmental influences such as altitude, precipitation, temperature, and soil (1).

As part of their experimental procedure, the researchers collected 371 GEFG samples from five Chinese provinces. Next, the research team used FT-IR spectroscopy to generate synchronous two-dimensional correlation spectroscopy (2D-COS) and three-dimensional correlation spectroscopy (3D-COS) images of the samples (1). Then, the researchers used a deep learning technique called residual neural network (ResNet) to achieve a 100% recognition accuracy when analyzing synchronous 2D-COS images (1).

Along with deep learning and FT-IR spectroscopy, the researchers employed the data-driven version of soft independent modeling of class analogy (DD-SIMCA) to distinguish between geographical indication (GI) production areas and non-GI areas. The authors have claimed the model has achieved 100% accuracy for GI production areas, ensuring that products from officially recognized regions could be authenticated with certainty (1). For non-GI areas, the specificity was 71.38%, indicating room for further refinement (1).

Several spectral regions are important for GEFG origin classification, which the study highlighted. For example, the 2D-COS images in the 1800-400 cm⁻¹ range were identified as a key spectral band (1).

The ability to rapidly and accurately verify GEFG’s geographic origin can help prevent mislabeling, protect regional branding, and support regulatory enforcement. Because GI-labeled herbal products are in high demand, they are often pricey. To ensure consumer demand remains high, it is integral for producers to ensure GEFG authenticity (1).

There are several key takeaways one could draw from this study. For one, combining FT-IR spectroscopy and deep learning models to discriminate herbs could be applied in studies concentrating on other plants and agricultural products. As this study showed, the integration of these two techniques proved to be effective. The second key takeaway is that further research is needed to refine the DD-SIMCA model’s specificity for non-GI areas (1). Expanding the sample size and incorporating additional environmental variables could improve its accuracy. Additionally, integrating this methodology into portable devices could make on-site geographic authentication a practical reality for farmers, traders, and regulators (1).

By integrating FT-IR spectroscopy with data-driven deep learning models, the researchers were able to identify geographic origins with remarkable precision, offering a promising tool for GI authentication. By setting a new benchmark in geographic traceability, this research paves the way for a more transparent and reliable herbal medicine market.

References

1. Deng, G.; Li, J.; Liu, H.; Wang, Y. Rapid Determination of Geographical Authenticity of Gastrodia elata f. glauca Using Fourier Transform Infrared Spectroscopy and Deep Learning. Food Cont. 2025, 167, 110810. DOI: 10.1016/j.foodcont.2024.110810
2. Yu, E.; Gao, Y.; Li, Y.; et al. An Exploration of Mechanism of High Quality and Yield of Gastrodia elata Bl. f. glauca by the Isolation, Identification and Evaluation of Armillaria. BMC Plant Biol. 2022, 22, 621. DOI: 10.1186/s12870-022-04007-8
3. Zheng, H.; Zhang, P.; Qin, J.; Guo, J.; Deng, J. High-throughput Sequencing-based Analysis of the Composition and Diversity of Endophytic Bacteria Community in Tubers of Gastrodia elata f. glauca. Front. Microbiol. 2023, 13, 1092552. DOI: 10.3389/fmicb.2022.109255