The data and findings discussed here come from the cited paper and do not represent test results from VISPEK equipment.
In the prediction of torreya protein, are shell scanning, shelling scanning and post-grinding scanning equally effective? This study allowed the same batch of samples to undergo different morphologies in sequence, directly comparing the changes brought about by the way they were presented. To understand the best model, you need to first confirm whether it corresponds to a complete nut or a broken particle.
Summary
Using Kjeldahl nitrogen as a reference, the study collected diffuse reflectance spectra from the same Chinese torreya nuts in three forms: in-shell, de-shelled with the seed coat retained, and ground particles. PLSR models were compared. Ground particles performed better after derivative and scatter correction, showing that sample preparation affects protein prediction. The evidence comes mainly from one origin; the ground-sample result cannot be presented as nondestructive in-shell detection performance.
(AI-generated conceptual illustration for context; it is not data from the paper, a photograph of the research equipment, or a VISPEK product diagram.)
Why It Matters
The comparison of the three states of the same nut shows that the protein detection plan should first determine the sample preparation method that the customer is willing to accept.
Methods and Reference Measurements
Mature Chinese torreya nuts from Zhejiang were scanned in-shell, then after shell removal with the seed coat intact, and finally after grinding and sieving into particles. A Smart Eye 1700 collected diffuse reflectance spectra at 1000–1650 nm with white-reference correction. Kjeldahl nitrogen provided the reference for the same homogenized material. PCA-MD and concentration residuals were used to identify outliers before comparing preprocessing and PLSR. The paired design links model differences to the way samples were presented.
Wavelengths and Acquisition Parameters
| Band/parameter | Original meaning and usage instructions |
|---|---|
| 1000–1650 nm | The range of portable spectral collection in this study |
| 1 nm／8 cm⁻¹ | The original article lists different resolution calibers at the same time, and there are doubts about the actual resolution expression. |
Key Findings
The particle sample 1Der-SNV-PLSR is the best, with corrected R²=0.92 and predicted R²=0.86; the particle results cannot be promoted as shell detection performance.
Limitations and Interpretation
The main evidence comes from single-origin harvest samples; grinding improves results meaning the best route is destructive.
Image copyright belongs to the paper’s original authors. License: CC BY 4.0.
Source Paper
Gu Y, Zhong H, Wu J, Li K, Huang Y, Fang H, Hassan M, Yao L, Zhao C. (2025). Application of Portable Near-Infrared Spectroscopy for Quantitative Prediction of Protein Content in Torreya grandis Kernels Under Different States. Foods (Basel, Switzerland). https://doi.org/10.3390/foods14111847
Contact Us
Whether portable equipment can be used for similar detection needs to be evaluated based on samples and experimental conditions, and the specific effect still needs to be verified through actual testing. Welcome to contact VISPEK to exchange your sample and testing needs, jointly evaluate experimental plans, and explore the feasibility of using our company's equipment for actual measurement and verification.
- Email: contact@vispek.cn
- Website: https://www.vispek.cn
VISPEK products shown for illustration; these were not the devices used in the cited study.