Posted in | News | Spectroscopy

Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi

In a recent article in Scientific Reports, researchers examined the use of Fourier Transform Near-Infrared (FT-NIR) spectroscopy with chemometric methods to more accurately identify the geographical origin of kimchi. With kimchi’s popularity increasing, this study looked at the issue of origin mislabeling, which can impact consumer trust and market trends.

Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi

Image Credit: krein1/Shutterstock.com

FT-NIR Spectroscopy: A Powerful Analytical Method

FT-NIR spectroscopy is a non-destructive technique that utilizes optics to analyze the chemical composition of samples, making it widely used for both qualitative and quantitative analysis in food science. By using the near-infrared region of the electromagnetic spectrum, this method can reveal a sample's chemical composition based on the interaction of near-infrared light with molecular vibrations, resulting in a spectrum that reflects the sample's chemical makeup.

A key advantage of FT-NIR spectroscopy is its ability to quickly provide data on chemical properties. This offers a practical alternative to traditional methods, which can be more time-consuming and less environmentally friendly.

Traditional techniques for identifying food origins, such as Principal Component Analysis (PCA), Hierarchical Cluster Analysis, and linear discriminant analysis (LDA), are costly and time-intensive. They also require skilled operators and can produce harmful waste.

In contrast, FT-NIR spectroscopy is faster, cost-effective, and more eco-friendly. When combined with chemometric tools, particularly machine learning algorithms like K-Nearest Neighbors (KNN), Support Vector Machines (SVM), and Partial Least Squares Discriminant Analysis (PLS-DA), FT-NIR can efficiently handle large datasets and accurately classify complex samples, even those with high similarity.

This combination offers an improvement in food authentication, which is particularly useful for products like kimchi, where origin can affect quality and consumer trust.

Using FT-NIR and Machine Learning for Investigating Kimchi Origin

In this paper, the authors developed a method for classifying the geographical origin of kimchi using FT-NIR spectroscopy. Rather than identifying specific marker compounds, they focused on interpreting spectral data directly with machine learning algorithms.

The study involved collecting and freeze-drying 60 kimchi samples (30 domestic and 30 imported). Spectral data was collected using an Antaris™ II FT-NIR analyzer, with each sample measured five times. Each measurement included 32 scans at a resolution of 8 wavenumbers within the range of 10,000-4,000 cm-1. Data acquisition was conducted under controlled conditions of 25 °C and 21-22 % relative humidity.

Spectral Data Preprocessing and Model Development

Before building the models, the researchers preprocessed the raw spectral data using TQ Analysis 9 software to improve model accuracy by reducing variations due to physical phenomena like scattering.

Three types of preprocessing methods were applied: scattering correction (using standard normal variate (SNV) and multiplicative signal correction (MSC)), spectral derivatives, and data smoothing for noise reduction. The choice of preprocessing methods could influence the classification task by either enhancing or diminishing relevant information, thereby affecting the performance of the machine learning algorithms.

Classification Model Performance

The authors evaluated several supervised machine learning algorithms, including K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Partial Least Squares Discriminant Analysis (PLS-DA), Random Forest (RF), Classification and Regression Trees (CART), and Naive Bayes (NB). Each model's performance was measured using metrics such as accuracy, recall, specificity, precision, and F1-score.

Key Findings and Insights

The outcomes showed that KNN and SVM achieved strong classification accuracy even without preprocessing, demonstrating their ability to handle complex spectral data. After applying appropriate preprocessing, RF and PLS-DA achieved perfect classification accuracy. However, CART and NB did not reach error-free classification, even with preprocessed data.

Among these methods, PLS-DA outperformed CART, NB, and RF on preprocessed datasets. KNN was particularly notable for its efficiency and ability to build a robust model without preprocessing, providing computational advantages.

While the FT-NIR approach did not highlight specific compounds linked to origin differences, it successfully distinguished domestic from imported kimchi samples based on their spectral fingerprints. The analysis also showed that the spectral data had notable similarities between domestic and imported samples, making full differentiation challenging.

Certain preprocessing methods, especially those involving derivatives and smoothing, effectively emphasized subtle differences in the spectral profiles. The study identified ten key peaks in the FT-NIR spectra associated with chemical composition differences, providing insights into potential biochemical markers of geographical origin.

Practical Applications

This research has important implications for food safety and consumer protection. Accurately identifying the geographical origin of kimchi can help prevent fraudulent labeling and support local producers by promoting authentic, domestically sourced products. This method may also be applicable to other fermented foods and agricultural products, improving traceability and authenticity throughout the food supply chain.

FT-NIR spectroscopy's nondestructive nature also allows for quick testing of large sample volumes, making it a useful tool for quality control in manufacturing and retail. Integrating machine learning adds to its analytical capabilities, enabling more detailed data analysis and interpretation in future studies.

Conclusion and Future Directions

In summary, FT-NIR spectroscopy combined with machine learning was an effective method for classifying kimchi by geographical origin. The findings highlight the potential for developing efficient, cost-effective, and environmentally friendly approaches to food authentication.

Future work could expand the dataset to include different types of kimchi and production conditions and explore this method’s application to other food products. Refining these techniques and broadening their use may support food labeling accuracy and consumer safety in the global market.

Journal Reference

Kim, SY., Ha, JH. (2024). Rapid determination of the geographical origin of kimchi by Fourier transform near-infrared spectroscopy coupled with chemometric techniques. Sci Rep. DOI: 10.1038/s41598-024-74662-4, https://www.nature.com/articles/s41598-024-74662-4

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.

Muhammad Osama

Written by

Muhammad Osama

Muhammad Osama is a full-time data analytics consultant and freelance technical writer based in Delhi, India. He specializes in transforming complex technical concepts into accessible content. He has a Bachelor of Technology in Mechanical Engineering with specialization in AI & Robotics from Galgotias University, India, and he has extensive experience in technical content writing, data science and analytics, and artificial intelligence.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Osama, Muhammad. (2024, November 13). Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi. AZoOptics. Retrieved on November 15, 2024 from https://www.azooptics.com/News.aspx?newsID=30046.

  • MLA

    Osama, Muhammad. "Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi". AZoOptics. 15 November 2024. <https://www.azooptics.com/News.aspx?newsID=30046>.

  • Chicago

    Osama, Muhammad. "Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi". AZoOptics. https://www.azooptics.com/News.aspx?newsID=30046. (accessed November 15, 2024).

  • Harvard

    Osama, Muhammad. 2024. Using FT-NIR and Machine Learning for Identifying Authenticate Kimchi. AZoOptics, viewed 15 November 2024, https://www.azooptics.com/News.aspx?newsID=30046.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.