Study On Rapid Detection Of Myoglobin Changes In Tan Mutton During Refrigerated Storage

Meat color is an important quality that affects the edible value and commercial value of Tan mutton,which directly affecting the decision for consumers to purchase.It changes with the amount and chemical state of myoglobin.The traditional detection methods of myoglobin are high cost,destructive,which require highly specialized equipment,highly trained people for the operation of instrument and preparation of samples.Therefore,some simple,sensitive,economical,reliable,and robust analytical instruments and methods are urgently needed to achieve accurate and fast non-destructive detection of myoglobin content in Tan mutton.This paper combined the ViS-NIR and NIR data of the Tan mutton with the chemometric method to explore the feasibility of rapid non-destructive detection of myoglobin content in Tan mutton by hyperspectral imaging technology and develop the quantitative function of myoglobin content,which provides a reference for the rapid determination of myoglobin content in Tan mutton.At the same time,the generalized two-dimensional correlation spectroscopy(G2D-COS)was used to explore the spectral changes of Tan mutton in order to provide an important scientific basis for the rapid non-destructive detection of myoglobin content of Tan mutton.The main research contents and conclusions are as follows:(1)Study on the change of myoglobin content in Tan mutton during storageWith the increase of storage time,the DeoMb content showed a decreased trend,MbO2 content increased first(p<0.05)and then decreased(p<0.05).The MetMb proportion increased with increasing storage time up to day 19(p<0.05),and then remained constant(p>0.05)until the end of the refrigerated storage time.(2)Study on non-destructive testing of the myoglobin content of Tan mutton using NIR hyperspectral imagingUsing NIR data,a prediction model of myoglobin content in mutton was established.The best model was optimized by comparing the modeling effects of 9 preprocessing methods and different feature wavelength extraction algorithms.The results showed that MA+UVE+VCPA-LSSVM model of DeoMb works well(R2c=0.832,RMSEC=1.124,R2p=0.808,RMSEP=1.170,RPD=2.290),the SG+iVISSA+VCPA-LSSVM model showed the best performance to determine MbO2(R2c=0.862,RMSEC=3.603,R2p=0.895,RMSEP=3.362,RPD=2.817),and the De-trending+G2D-COS-LSSVM model works well in predicting MetMb(R2c=0.866,RMSEC=3.555,R2p=0.879,RMSEP=3.507,RPD=2.835).At the same time,using the G2D-COS method,four main correlation spectral changes caused by fluctuations in DeoMb content occurred in the sequence is 1080 nm-1055 nm-1243 nm-1318 nm.Three main correlation spectral changes caused by fluctuations in MbO2 content occurred in the sequence is 1120 nm-1061 nm-1252 nm.Four main correlation spectral changes caused by fluctuations in DeoMb content occurred in the sequence is 1037 nm-1151 nm-1620 nm-1446 nm.The results indicated that 2nd N-H stretching of NH in amides changes firstly,then the 2nd C-H stretching of CH2 groups,1st O-H stretching of COOH groups,and finally the absorption of C=O stretching in the amide I band of proteins.This method provided significant spectral resolution,which could distinguish bands arising from different sources,as well as the order of intensity changes.(3)Study on non-destructive testing of the myoglobin content of Tan mutton using Vis-NIR hyperspectral imagingThe myoglobin content of the sample was measured by a spectrophotometer,and the region of interest of the 200 sample spectral images during storage were extracted by ENVI4.8 software.Nine spectral preprocessing techniques such as SNV、MSC、Baseline、De-treding、SG、2nd derivative、MA、GF、MF were applied to eliminate noise.CARS、iVISSA、SPA、VCPA、UVE、IRF、G2D-COS were used to select and optimize variables.PLSR and LSSVM mothod were used to develope prediction models based on full-band and feature bands.The results showed that the CARS-LSSVM model works well in predicting DeoMb(R2c=0.891,RMSEC=0.897,R2p=0.810,RMSEP=1.127,RPD=2.377),the SNV+iVISSA+CARS-LSSVM model showed the best performance to determine MbO2(R2c=0.902,R2p=0.914,RPD=3.644),and the De-trending+iVISSA+CARS-LSSVM model works well in predicting MetMb(R2c=0.908,RMSEC=2.786,R2p=0.915,RMSEP=2.777,.RPD=3.591).At the same time,using the G2D-COS method,five main correlation spectral changes caused by fluctuations in DeoMb content occurred in the sequence is 458 nm-492 nm-588 nm-612 nm-982 nm-747 nm.Five main correlation spectral changes caused by fluctuations in MbO2 content occurred in the sequence is 574 nm-593 nm-454 nm-973 nm-752 nm.Six main correlation spectral changes caused by fluctuations in DeoMb content occurred in the sequence is 430 nm-425 nm-574 nm-612 nm-780 nm-973 nm.The characteristic sequences of spectral intensity changes of the main relevant bands of the three myoglobins were almost the same or similar,as the myoglobin content increased,the Myoglobin-related band intensity changed before the second and third overtones of O-H stretching generated by water bands.However,the positions of autopeaks have shifted slightly.At the same time,comparing and analyzing the model effects of Vis-NIR and NIR regions,it could be clearly found that the modeling effect of Vis-NIR is significantly better than the model effect of NIR regions,which indicating that the 400-1000 nm band range is more suitable for detecting myoglobin content.The overall results from this study indicated that it was feasible to predict myoglobin contents in Tan sheep using HSI.