Differentiation between fresh and frozen-thawed mackerel fish using low-cost portable near infrared spectrometry devices

Giró-Candanedo, Mar; Cruz, Jordi; Comaposada, Josep; Barnés-Calle, Clara; Gou, Pere; Fulladosa, Elena

doi:10.1016/j.jfoodeng.2024.112131

Cited by 2 publications

(2 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, a small shift in the water band at 1416 nm to a higher wavelength was found for frozen samples. A similar shift in water bands in the region between 990 and 1060 nm was reported by Giró-Candanedo et al [ 42 ] for fresh and frozen–thawed mackerel fish. There were small differences between spectra of the same sample measured after the first and second freezing–thawing cycles.…”

Section: Resultssupporting

confidence: 87%

“…Precision higher than 88% for the differentiation of fresh and defrosted common white fish species was reported by Goncalves et al [ 25 ] using a handheld NIR spectrometer in the same spectral range of 900–1700 nm. Giró-Candanedo et al [ 42 ] investigated the discriminative power of low-cost NIR devices and PLS-DA in classification models for fresh and thawed fish samples with a rate for fresh fillets of 90.3% and 94.1% and for frozen samples of 91.2% and 89.7%, respectively. Kaltenbach et al [ 20 ], on the base of the 1 H NMR analysis of the lipid fraction of fish and PLS-DA, differentiated fresh from frozen–thawed fish and reported a correct prediction of 90%.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Near-Infrared Spectroscopy for Rapid Differentiation of Fresh and Frozen–Thawed Common Carp (Cyprinus carpio)

Atanassova,

Yorgov,

Stratev

et al. 2024

Sensors

View full text Add to dashboard Cite

This study aimed to investigate near-infrared spectroscopy (NIRS) in combination with classification methods for the discrimination of fresh and once- or twice-freeze–thawed fish. An experiment was carried out with common carp (Cyprinus carpio). From each fish, test pieces were cut from the dorsal and ventral regions and measured from the skin side as fresh, after single freezing at minus 18 °C for 15 ÷ 28 days and 15 ÷ 21 days for the second freezing after the freeze–thawing cycle. NIRS measurements were performed via a NIRQuest 512 spectrometer at the region of 900–1700 nm in Reflection mode. The Pirouette 4.5 software was used for data processing. SIMCA and PLS-DA models were developed for classification, and their performance was estimated using the F1 score and total accuracy. The predictive power of each model was evaluated for fish samples in the fresh, single-freezing, and second-freezing classes. Additionally, aquagrams were calculated. Differences in the spectra between fresh and frozen samples were observed. They might be assigned mainly to the O–H and N–H bands. The aquagrams confirmed changes in water organization in the fish samples due to freezing–thawing. The total accuracy of the SIMCA models for the dorsal samples was 98.23% for the calibration set and 90.55% for the validation set. For the ventral samples, respective values were 99.28 and 79.70%. Similar accuracy was found for the PLS-PA models. The NIR spectroscopy and tested classification methods have a potential for nondestructively discriminating fresh from frozen–thawed fish in as methods to protect against fish meat food fraud.

show abstract

Section: Resultssupporting

confidence: 87%