Skip to main content
NCBI home page
Journal of Food Science and Technology logoLink to Journal of Food Science and Technology
. 2023 Nov 10;61(5):950–957. doi: 10.1007/s13197-023-05890-1

Identification of individual goat animals by means of near infrared spectroscopy and chemometrics analysis of commercial meat cuts

D Cozzolino 1,2,✉, S Zhang 1,2, A Khole 1,2, Z Yang 1,2, P Ingle 1,2, M Beya 1,2, P F van Jaarsveld 1,2, D Bureš 3,4, L C Hoffman 1,2
PMCID: PMC10933230  PMID: 38487278

Abstract

Although the identification of animal species and muscles have been reported previously, no studies have been found on the use of NIR spectroscopy to identify individual animals from the analysis of commercial meat cuts. The aim of this study was to evaluate the use of a portable near infrared (NIR) instrument combined with classical chemometrics methods [principal component analysis (PCA) and partial least squares discriminant analysis PLS-DA)] to identify the origin of individual goat animals using the spectral signature of their commercial cut. Samples were collected from several carcasses (6 commercial cuts x 24 animals) sourced from a commercial abattoir in Queensland (Australia). The NIR spectra of the samples were collected using a portable NIR instrument in the wavelength range between 950 and 1600 nm. Overall, the PLS-DA models correctly classify 82% and 79% of the individual goat samples using either the goat rack or loin cut samples, respectively. The study demonstrated that NIR spectroscopy was able to identify individual goat animals based on the spectra properties of some of the commercial cut samples analysed (e.g. loin and rack). These results showed the potential of this technique to identify individual animals as an alternative to other laboratory methods and techniques commonly used in meat traceability.

Keywords: Goat, Commercial cut, NIR, Chemometrics, Traceability, Provenance

Introduction

Meat export countries need to maintain their reputation for producing safe and high-quality products, where effective analytical and monitoring systems are required for this task (e.g. authenticity and traceability systems) (Manning and Soon 2016; Olsen and Borit 2018). Consumers also want to know more about the food they buy, and demand for information about the nutrition, safety and origin of the foods (e.g. provenance, contamination), biosecurity status (e.g. diseases and pests), as well as other issues such as sustainability and animal welfare (Olsen and Borit 2018; The Traceability Working Group 2019). Monitoring authenticity and provenance in the meat supply and value chains will improve transparency and strengthening assurances that will be translated to build consumer trust about meat and meat products (Primrose et al. 2010; Cawthorn et al. 2013). To achieve this, several analytical techniques, methods and even systems have been proposed, where the use of instrumental techniques such as near infrared (NIR) spectroscopy has been reported by different researchers (Prieto et al. 2009, 2017).

In recent years, animal identification and traceability have become very important not only to assure consumer confidence in meat consumption, but also as a competitive advantage to countries where meat is exported (Zhao et al. 2020). Several techniques are available in the market to trace individual animals, including the utilization of radio frequency identification (RFID) tags in live animals, DNA fingerprinting and the use of stable isotopes (Zhao et al. 2020). Some of these techniques have the highest analytical accuracy, but the complexity of the method or technique used combined with the high cost may limit its application by the industry (Manning and Soon 2016; Olsen and Borit 2018).

Techniques that belong to the field of vibrational spectroscopy, are characterised for the measurement of the absorption, emission or scattering of light by organic molecules present in the meat at specific wavelengths (Cozzolino 2012, 2015; Roberts and Cozzolino 2016). The amount of light absorbed by these molecules is associated with a characteristic chemical bond (e.g. C–H, N–H, OH) as well as its concentration in the sample (Cozzolino 2015). In practice, when molecules are exposed to infrared (IR) radiation, these molecules undergo rotational, vibrational, electronic, or ionization processes, in order of increasing energy (Cozzolino 2015; Roberts and Cozzolino 2016). Specifically in the IR region, once the sample is radiated with light the different chemical bonds (e.g. C–H, N–H) absorb the energy at different wavelengths, depending on the atoms and bonds, the surrounding molecules, or the type of vibration (Cozzolino 2015; Roberts and Cozzolino 2016). The energy from the different bonds is then collected and the individual spectrum of a given sample is then generated. Consequently, the IR spectrum of a sample record the chemical, physical and functional properties of a sample (Cozzolino 2015; Roberts and Cozzolino 2016). Overall, wavelengths in the UV-visible (VIS), MIR, and NIR range have been reported and utilised in a wide range of application in the meat industry to measure composition as well as to monitor issues associated with meat authenticity, fraud, provenance and traceability (Cozzolino 2015; Roberts and Cozzolino 2016; Prieto et al. 2009, 2017).

In authentication, fraud, provenance and traceability applications, the IR spectrum or trace is then analysed using different chemometric or multivariate data techniques such as principal component (PCA), discriminant (DA) and cluster (CA) analysis (Cozzolino 2012, 2021; Power and Cozzolino 2020; Roberts and Cozzolino 2016). In practice, after the IR spectrum is collected and recorded, patterns or groups of samples can be differentiated having similar or different characteristics or properties using chemometric methods and techniques (Cozzolino 2012, 2021; Power and Cozzolino 2020; Roberts and Cozzolino 2016).

The application of NIR spectroscopy in the field of meat sciences is not new (Prieto et al. 2009, 2017). Several reports demonstrated the use of NIR spectroscopy to measure the proximate composition (e.g. protein, fat and moisture content) in both individual muscles and commercial cut samples (Maduro Dias et al. 2021), to measure the content of intramuscular fat (IMF) and pH in either animal carcasses or individual muscles in both domesticated and wild animal species (Prieto et al. 2009, 2017; Dixit et al. 2017; Chapman et al. 2020).

In addition to the prediction of proximate composition, NIR spectroscopy has been extensively used for its ability to evaluate a wide range of issues associated with meat fraud and provenance (López-Maestresalas et al. 2019), the identification and discrimination of animal species (Alamprese et al. 2013, 2016) and individual muscles (Cozzolino and Murray 2004; Dumalisile et al. 2020), and in a wide range of applications related with the identification of adulteration issues in different meat species (e.g. identification of animal species in meat mixtures) (Cozzolino 2016; Bai et al. 2021; Edwards et al. 2020, 2021). The use of NIR spectroscopy in this field has proved that individual muscles and species can be detected with results comparable to available diagnostic and laboratory techniques such as DNA and isotope analysis (Cawthorn et al. 2013; Edwards et al. 2020, 2021; Mamani-Linares et al. 2012). Recently, developments in portable and hand-held instrumentation have also boosted innovative applications of NIR spectroscopy in the field of meat sciences (Kademi et al. 2019; Kucha and Ngadi 2020; Savoia et al. 2020).

More recently the ability of hyperspectral imaging (HYPER) in the visible (VIS) and NIR ranges to analyse meat composition has been also demonstrated by different authors (Bai et al. 2021; Jia et al. 2022). Applications of HYPER imaging has been reported to analyse the proximate composition and tenderness of meat (León-Ecay et al. 2022) as well as in the authentication and identification of fraud in different species of meat (Barbin et al. 2012; Kamruzzaman et al. 2013, 2015; Jiang et al. 2021; Ropodi et al. 2015; Zhang et al. 2022; Zheng et al. 2019).

Although the identification of animal species and muscles have been reported, no studies have been found on the use of NIR spectroscopy to identify individual animals from the analysis of a muscle or commercial cut from a given animal. The possibility to identify an individual animal using NIR spectra of a commercial cut can open new possibilities in animal traceability as well as providing with new tools to either monitor meat provenance or to detect fraud along the meat supply and value chains.

The aim of this study was to evaluate the use of a portable NIR instrument combined with classical chemometrics methods (principal component analysis and partial least squares discriminant analysis) to identify the origin of an individual animal using the spectral signature of their commercial cuts.

Materials and methods

Sample collection

The commercial cut samples were obtained from chilled carcasses after 24 h slaughter (Hoffman et al. 2022). Sample characteristics including age, sex, carcass weight, experimental conditions and other relevant information can be found in a previous report (Hoffman et al. 2022). In this study, six fresh commercial meat cuts namely back leg, chump or leg, flap or breast, loin, rack, and shoulder were obtained from individual goat carcasses (n = 24) sourced from a commercial abattoir in Queensland (Australia). The location of the commercial cut samples in the carcass of the goat is illustrated in Fig. 1. The origin of the samples was verified using the data provided by the producer to the commercial abattoir.

Fig. 1.

Fig. 1

Open in a new tab

Location of the commercial cut samples in the carcass of the goat

Spectra collection

The NIR spectra of the commercial meat cut samples were collected using a portable NIR instrument (Micro-NIR 1700. Viavi, Milpitas, California, USA) operating in the wavelength range between 950 and 1600 nm (10 nm wavelength resolution). The spectra collection and instrument set up was controlled using the proprietary software provided by the instrument manufacturer (MicroNIR Prov 3.1, Viavi, Milpitas, California, USA). The spectral data acquisition parameters were set at 50 ms integration time and averaging of 50 scans (MicroNIR Prov 3.1, Viavi, Milpitas, CA, USA). Every 20 samples, a reference spectrum was collected using a Spectralon® tile. For each commercial cut sample, three scans were collected at different random positions within the sample. Therefore, the total number of samples collected and used in this study were n = 432 (24 animals x 6 commercial meat cuts x 3 replicates).

Statistical analysis

The NIR data was exported in Excel format (*.xls) into The Unscrambler software (version X, CAMO, Oslo, Norway) for data analysis and pre-processing. The NIR spectra was pre-processed using the Savitzky-Golay second derivative (21 smoothing points and second polynomial order) prior to spectra interpretation and chemometric analysis (Saviztky and Golay 1964). Principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were used to analyse the data as well as to develop classification models based on the commercial cut samples. The PLS-DA models were developed using the individual animal or carcass number as dummy variable (e.g. carcass 1 or animal 1, carcass 2 or animal 2, etc.). All models were developed and validated using full cross validation (leave one out) (Bureau et al. 2019; Næs et al.2002; Williams et al. 2017). The models were assessed using the coefficient of determination in cross validation (R2), the standard error in cross validation (SECV), the bias, and slope (Bureau et al. 2019; Næs et al. 2002). The per cent of correct classification (CC) was also used to assess the performance of the classification models developed.

Results and discussion

The PCA score plot developed using all samples (all commercial cut and goat samples) did not show any clear separation between them (Fig. 2). This can be explained by the complexity of the data, associated with the different commercial cut and carcass samples (individual animals) analysed. Therefore, it was decided to build separate PCA’s using each of the commercial meat cut samples collected from all the goat animals. Each PCA developed (score plots and loadings) using the individual commercial cut samples showed different trends or patterns. The PCA score plot developed using only the rack samples showed the best separation among the individual goat animals (Fig. 3), compared with the other commercial cut samples analysed (data not shown).

Fig. 2.

Fig. 2

Open in a new tab

Principal component score plot obtained from the use all individual goat and commercial cut samples analysed using near infrared reflectance spectroscopy

Fig. 3.

Fig. 3

Open in a new tab

Principal component score plot obtained from the goat rack samples analysed using near infrared reflectance spectroscopy. Please note that same number indicates an individual goat animal

The PCA score plot using the rack samples showed that 97% of the total variance in the data set is explained by the first three principal components (PC) (PC1 = 57%, PC2 = 34% and PC3 = 6%). The highest loadings explaining the separation between samples were observed in wavelengths around 976 nm (O–H) (observed in all PCs) associated with water or moisture content, around 1162 nm (C–H) (observed only in PC1 and PC2) associated with lipids, around 1131 nm (C-H) (observed only in PC3), 1242 nm (C–H) (observed only in PC3), and 1372 nm (C–H) (observed only in PC3) associated with fat or lipid content and between 1409 and 1416 nm (O–H) associated with water and protein content (Fig. 4) (Workman and Weyer 2008; Cozzolino and Murray 2004).

Fig. 4.

Fig. 4

Open in a new tab

Loadings derived from the principal component analysis of the goat rack samples analysed using near infrared reflectance spectroscopy

In addition to the PCA analysis, PLS-DA models were developed using the combination of all animals and commercial cut samples (back leg, chump, flap, loin, rack, shoulder). The cross-validation statistics for the prediction of single animals using all samples and commercial cut samples yield a R2 of 0.72 and SECV of 2.1 (see Table 1). In this case an R2 of 0.72 indicated that 72% of the variance in the data set can be explained using NIR spectroscopy. The cross-validation statistics for the prediction of a single animal using individual commercial cut samples gave a R2 of 0.59 (SECV = 2.57); R2 of 0.62 (SECV = 2.5), and R2 of 0.73 (SECV = 2.1) using the loin, chump, and rack samples, respectively. The R2 indicated that 59%, 62% and 73% of the variance in the data can be explained by NIR spectroscopy for the measurement of loin, chump and rack samples, respectively. No satisfactory PLS-DA models were obtaining using the other commercial cut samples analysed. The observed differences in the cross-validation statistics can be associated with the characteristics of either the individual or the combination of muscles that constitute the commercial cut sample. For example, both rack and loin are predominantly made up of one muscle while the other commercial cut samples analysed are made up from the combination of different muscles having different physiological functions or properties. In addition, other characteristics or properties that can explain the differences in some of the cut samples are the breed, gender, age, and environment conditions. Overall, the PLS-DA models correctly classify 82% and 79% of the individual goat samples using the goat rack and loin cut, respectively.

Table 1.

Cross validation statistics for the prediction of individual animals using all samples as well as individual commercial cut samples analysed using near infrared reflectance spectroscopy

R2 SECV Slope Bias LV
Loin 0.59 2.6 0.67 − 0.001 11
Rack 0.72 2.1 0.77 0.02 11
Chump 0.62 2.5 0.70 − 0.004 12
All samples 0.72 2.1 0.71 0.001 14
Open in a new tab

R2: coefficient of determination in cross validation; SECV: standard error in cross validation; LV: number of latent variables used by the model

The PLS-DA loadings were also interpreted as they provide with information about the main wavelengths used by the model (Fig. 5). Each of the cross-validated models has different and unique loadings, although the number of latent variables (LV) used were similar (between 11 and 14). The main differences between loadings were observed in wavelengths around 1050 nm (rack), 1137 nm (loin), at 1236 nm (chump) and around 1340 nm (chump and rack), 1409 nm (loin) and 1384 and 1437 nm (loin). These wavelengths correspond with the presence of O-H and C-H bonds associated with lipids, protein and moisture content of the commercial cut samples analysed as described above (Workman and Weyer 2008; Cozzolino and Murray 2004). These results confirm that each model is different due to the intrinsic characteristics and properties of the commercial cut samples analysed (e.g. type and number of muscles, physiological properties and characteristics).

Fig. 5.

Fig. 5

Open in a new tab

Loadings derived from the partial least squares discriminant analysis used to classify goat chump, loin and rack samples analysed using near infrared reflectance spectroscopy

Although the use of NIR spectroscopy alone does not guarantee the proper identification of individual animals (e.g. 100% correct classification), the appropriate application and implementation of this technique can provide with the information required to develop tools for quality assurance and traceability of meat throughout the supply and value chains. Ideally, NIR spectroscopy should be combined with other techniques such as RFID tags and DNA fingerprint to trace the meat of individual cuts along the supply and value chains.

Conclusion

This study demonstrated that a portable NIR instrument combined with chemometrics can identify individual goat animals based on the spectra properties. Loadings derived from the PCA and PLS-DA analysis were observed to be different due to the chemical properties of the commercial cut samples analysed. The PLS-DA models indicated that 82% and 79% of the individual goat samples can be correctly classified as rack and loin cut, respectively. Overall, these results showed the potential of NIR spectroscopy to separate individual animals based on the information collected from some of the commercial cut samples (e.g. rack and loin). In the future, we can expect an improvement in the NIR cross-validation models by incorporating samples from other commercial and production conditions, as well as different genetics, age, production systems.

Author contributions

HLC, IP, HKA, ZS, YZ; BM, vJPF, BD, and CD: data collection and analysis; HLC, BD, and CD: draft preparation, and editing of manuscript; HLC. and DC: supervision. HLC: project administration. All authors have read and agreed to the published version of the manuscript.

Funding

Internal University of Queensland Institutional funds. The support of the Ministry of Agriculture of the Czech Republic (MZE-RO0718) is also acknowledged.

Data availability

Upon request.

Code availability

Not applicable.

Declarations

Conflict of interest

The authors declare no conflict of interest.

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Alamprese C, Casale M, Sinelli N, Lanteri S, Casiraghi E. Detection of minced beef adulteration with Turkey meat by UV–vis, NIR and MIR spectroscopy. LWT - Food Sci Technol. 2013;53:225–232. doi: 10.1016/j.lwt.2013.01.027. [DOI] [Google Scholar]
  2. Alamprese C, Amigo JM, Casiraghi E, Engelsen SB. Identification and quantification of Turkey meat adulteration in fresh, frozen-thawed and cooked minced beef by FT-NIR spectroscopy and chemometrics. Meat Sci. 2016;121:175–181. doi: 10.1016/j.meatsci.2016.06.018. [DOI] [PubMed] [Google Scholar]
  3. Bai Y, Liu H, Zhang B, Zhang J, Wu H, Zhao S, Qie M, Guo J, Wang Q, Zhao Y. Research progress on traceability and authenticity of beef. Food Rev Int. 2021 doi: 10.1080/87559129.2021.1936000. [DOI] [Google Scholar]
  4. Barbin D, Elmasry G, Sun DW, Allen P. Near-infrared hyperspectral imaging for grading and classification of pork. Meat Sci. 2012;90:259–268. doi: 10.1016/j.meatsci.2011.07.011. [DOI] [PubMed] [Google Scholar]
  5. Bureau S, Cozzolino D, Clark CJ. Contributions of Fourier-transform mid infrared (FT-MIR) spectroscopy to the study of fruit and vegetables: a review. Post Bio Technol. 2019;148:1–14. doi: 10.1016/j.postharvbio.2018.10.003. [DOI] [Google Scholar]
  6. Cawthorn D-M, Steinman HA, Hoffman LC. A high incidence of species substitution and mislabelling detected in meat products sold in South Africa. Food Control. 2013;32:440–449. doi: 10.1016/j.foodcont.2013.01.008. [DOI] [Google Scholar]
  7. Chapman J, Elbourne A, Truong VK, Cozzolino D. Shining light into meat—a review on the recent advances in in vivo and carcass applications of near infrared spectroscopy. Int J Food Sci Technol. 2020;55:935–941. doi: 10.1111/ijfs.14367. [DOI] [Google Scholar]
  8. Cozzolino D. Recent trends on the use of infrared spectroscopy to trace and authenticate natural and agricultural food products. App Spectros Rev. 2012;47:518–530. doi: 10.1080/05704928.2012.667858. [DOI] [Google Scholar]
  9. Cozzolino D. The role of vibrational spectroscopy as tool to assess economical motivated Fraud and counterfeit issues in agricultural products and foods. Anal Methods. 2015;7:9390–9400. doi: 10.1039/C5AY01792K. [DOI] [Google Scholar]
  10. Cozzolino D (2016) Near infrared spectroscopy and food authenticity. In: Advances in food traceability techniques and technologies, pp. 119–136. Elsevier. 10.1016/B978-0-08-100310-7.00007-7
  11. Cozzolino D. Editorial special issue food traceability and security. Int J Food Sci Technol. 2021;56:2579–2579. doi: 10.1111/ijfs.15050. [DOI] [Google Scholar]
  12. Cozzolino D, Murray I. Identification of animal meat muscles by visible and near infrared reflectance spectroscopy. LWT-Food Sci Technol. 2004;37:447–452. doi: 10.1016/j.lwt.2003.10.013. [DOI] [Google Scholar]
  13. Dixit Y, Casado-Gavalda MP, Cama-Moncunill R, Cullen PJ, Sullivan C. Challenges in model development for meat composition using multipoint NIR spectroscopy from at-line to in-line monitoring: multipoint NIR spectroscopy. J Food Sci. 2017;82:1557–1562. doi: 10.1111/1750-3841.13770. [DOI] [PubMed] [Google Scholar]
  14. Dumalisile P, Manley M, Hoffman LC, Williams PJ. Discriminating muscle type of selected game species using near infrared (NIR) spectroscopy. Food Control. 2020;110:106981. doi: 10.1016/j.foodcont.2019.106981. [DOI] [Google Scholar]
  15. Edwards K, Manley M, Hoffman LC, Beganovic A, Kirchler CG, Huck CW, Williams PJ. Differentiation of South African game meat using near-infrared (NIR) spectroscopy and hierarchical modelling. Molecules. 2020;25:1845. doi: 10.3390/molecules25081845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Edwards K, Manley M, Hoffman LC, Williams PJ. Non-destructive spectroscopic and imaging techniques for the detection of processed meat Fraud. Foods. 2021;10:448. doi: 10.3390/foods10020448. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hoffman LC, Ingle P, Khole AH, Zhang S, Yang Z, Beya M, Bureš D, Cozzolino D. Characterisation and identification of individual intact goat muscle samples (Capra sp.) using a portable near-infrared spectrometer and chemometrics. Foods. 2022;11:2894. doi: 10.3390/foods11182894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jia W, van Ruth S, Scollan N, Koidis A. Hyperspectral imaging (HSI) for meat quality evaluation across the supply chain: current and future trends. Curr Res Food Sci. 2022;5:1017–1027. doi: 10.1016/j.crfs.2022.05.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Jiang H, Ru Y, Chen Q, Wang J, Xu L. Near-infrared hyperspectral imaging for detection and visualization of offal adulteration in ground pork. Spectrochim Acta Part A Mol Biomol Spectros. 2021;249:119307. doi: 10.1016/j.saa.2020.119307. [DOI] [PubMed] [Google Scholar]
  20. Kademi HI, Ulusoy BH, Hecer C. Applications of miniaturized and portable near infrared spectroscopy (NIRS) for inspection and control of meat and meat products. Food Rev Int. 2019;35:201–220. doi: 10.1080/87559129.2018.1514624. [DOI] [Google Scholar]
  21. Kamruzzaman M, Sun DW, ElMasry G, Allen P. Fast detection and visualization of minced lamb meat adulteration using NIR hyperspectral imaging and multivariate image analysis. Talanta. 2013;103:130–136. doi: 10.1016/j.talanta.2012.10.020. [DOI] [PubMed] [Google Scholar]
  22. Kamruzzaman M, Makino Y, Oshita S, Liu S. Assessment of visible near-infrared hyperspectral imaging as a tool for detection of horsemeat adulteration in minced beef. Food Bioprocess Technol. 2015;8:1054–1062. doi: 10.1007/s11947-015-1470-7. [DOI] [Google Scholar]
  23. Kucha CT, Ngadi MO. Rapid assessment of pork freshness using miniaturized NIR spectroscopy. J Food Measur Charact. 2020;14:1105–1115. doi: 10.1007/s11694-019-00360-9. [DOI] [Google Scholar]
  24. León-Ecay S, López-Maestresalas A, Murillo-Arbizu MT, Beriain MJ, Mendizabal JA, Arazuri S, Jarén C, Bass PD, Colle MJ, García D. Classification of beef longissimus thoracis muscle tenderness using hyperspectral imaging and chemometrics. Foods. 2022;11:3105. doi: 10.3390/foods11193105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. López-Maestresalas A, Insausti K, Jarén C, Pérez-Roncal C, Urrutia O, Beriain MJ, Arazuri S. Detection of minced lamb and beef Fraud using NIR spectroscopy. Food Control. 2019;98:465–473. doi: 10.1016/j.foodcont.2018.12.003. [DOI] [Google Scholar]
  26. Maduro Dias CSAM, Nunes HP, Melo TMMV, Rosa HJD, Silva CCG, Borba AES. Application of near infrared reflectance (NIR) spectroscopy to predict the moisture, protein, and fat content of beef for gourmet hamburger preparation. Liv Sci. 2021;254:104772. doi: 10.1016/j.livsci.2021.104772. [DOI] [Google Scholar]
  27. Mamani-Linares LW, Gallo C, Alomar D. Identification of cattle, llama and horse meat by near infrared reflectance or transflectance spectroscopy. Meat Sci. 2012;90:378–385. doi: 10.1016/j.meatsci.2011.08.002. [DOI] [PubMed] [Google Scholar]
  28. Manning L, Soon JM. Food Safety, Food Fraud, and Food Defense: a fast evolving literature. J Food Sci. 2016;81:823–834. doi: 10.1111/1750-3841.13256. [DOI] [PubMed] [Google Scholar]
  29. Næs T, Isaksson T, Fearn T, Davies T (2002) A user-friendly guide to multivariate calibration and classification, 2nd edn, vol 6. NIR Chichester. 10.1255/978-1-906715-25-0
  30. Olsen O, Borit M. The components of a food traceability system. Trends Food Sci Technol. 2018;77:143–149. doi: 10.1016/j.tifs.2018.05.004. [DOI] [Google Scholar]
  31. Power A, Cozzolino D. How fishy is your fish? Authentication, provenance and traceability in fish and seafood by means of vibrational spectroscopy. Appl Sci. 2020;10:4150. doi: 10.3390/app10124150. [DOI] [Google Scholar]
  32. Prieto N, Roehe R, Lavin P, Batten G, Andres S. Application of near infrared reflectance spectroscopy to predict meat and meat products quality: a review. Meat Sci. 2009;83:175–186. doi: 10.1016/j.meatsci.2009.04.016. [DOI] [PubMed] [Google Scholar]
  33. Prieto N, Pawluczyk O, Edward M, Dugan R, Aalhus JL. A review of the principles and applications of near-infrared spectroscopy to characterize meat, fat, and meat products. Appl Spectros. 2017;7:1406–1426. doi: 10.1177/0003702817709299. [DOI] [PubMed] [Google Scholar]
  34. Primrose S, Woolfe M, Rollinson S. Food forensics: methods for determining the authenticity of foodstuffs. Trends Food Sci Technol. 2010;21(12):582–590. doi: 10.1016/j.tifs.2010.09.006. [DOI] [Google Scholar]
  35. Roberts JJ, Cozzolino D. An overview on the application of chemometrics in food science and technology - an approach to quantitative data analysis. Food Anal Methods. 2016;9:3258–3267. doi: 10.1007/s12161-016-0574-7. [DOI] [Google Scholar]
  36. Ropodi AI, Pavlidis DE, Mohareb F, Panagou EZ, Nychas GJE. Multispectral image analysis approach to detect adulteration of beef and pork in raw meats. Food Res Int. 2015;67:12–18. doi: 10.1016/j.foodres.2014.10.032. [DOI] [Google Scholar]
  37. Saviztky A, Golay MJE. Smoothing and differentiation of data by simplified least squares procedures. Anal Chem. 1964;36:1627–1639. doi: 10.1021/ac60214a047. [DOI] [Google Scholar]
  38. Savoia S, Albera A, Brugiapaglia A, Di Stasio L, Ferragina A, Cecchinato A, Bittante G. Prediction of meat quality traits in the abattoir using portable and hand-held near-infrared spectrometers. Meat Sci. 2020;161:108017. doi: 10.1016/j.meatsci.2019.108017. [DOI] [PubMed] [Google Scholar]
  39. The Traceability Working Group (2019) National Traceability Framework: enhancing Australia’s world-class agricultural traceability systems, Canberra, Australia
  40. Williams P, Dardenne P, Flinn P. Tutorial: items to be include in a report on a near infrared spectroscopy project. J Near Infrared Spectros. 2017;25:85–90. doi: 10.1177/0967033517702395. [DOI] [Google Scholar]
  41. Workman J, Weyer L. Practical guide to interpretive near-infrared spectroscopy. Boca Raton: CRC Press Taylor and Francis Group; 2008. [Google Scholar]
  42. Zhang Y, Zheng M, Zhu R, Ma R. Detection of adulteration in mutton using digital images in time domain combined with deep learning algorithm. Meat Sci. 2022;192:108850. doi: 10.1016/j.meatsci.2022.108850. [DOI] [PubMed] [Google Scholar]
  43. Zhao J, Li A, Jin X, Pan L. Technologies in individual animal identification and meat products traceability. Biotech Biotechnol Equip. 2020;34:48–57. doi: 10.1080/13102818.2019.1711185. [DOI] [Google Scholar]
  44. Zheng X, Li Y, Wei W, Peng Y. Detection of adulteration with duck meat in minced lamb meat by using visible near-infrared hyperspectral imaging. Meat Sci. 2019;149:55–62. doi: 10.1016/j.meatsci.2018.11.005. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Upon request.

Not applicable.

Articles from Journal of Food Science and Technology are provided here courtesy of Springer

RESOURCES