Raman spectroscopy for estimation of residual bone minerals in mechanically deboned chicken meat

Residual bone, measured as %calcium or %ash, is a strictly controlled quality parameter of mechanically deboned chicken meat (MDCM). Raman spectroscopy was developed as a rapid tool for estimation of this important parameter. 

Mechanical deboning is a process where protein-reach mince is optimally recovered from animal carcasses. This process typically involves grinding of carcasses to create a meat-bone slurry and subsequently pass the slurry through a thieving system to separate out the meat from bone residues. According to European food safety authorities, one of the control parameters for mechanically separated meat is calcium (or ash) content, which is an indicator of residual bone. Currently there is no analytical tool that allows the rapid measurement of both calcium or ash levels in meat and bone mixtures in industrial environment.  

Bone Granulas

Depending on the process settings and carcass composition, fine granules of bone could be introduced to MDCM. Therefore, bone content of such meat is usually controlled by setting calcium or ash limits. The currently practiced analytical procedures for determination of calcium and ash in mechanically separated meat are based on methods such as titration and combustion, respectively.  

These methods are time consuming and are typically performed offline on few grams of sample that are expected to represent an entire batch of production. Therefore, these methods cannot be directly used to control calcium and ash content in a large-scale industrial production. 

Raman Spectroscopy

In the present study we have developed Raman spectroscopy as a rapid tool for estimation residual bone (i.e., ash and calcium) in MDCM. In contrast to the existing methods for measuring calcium content, e.g. titration, the presented strategy is rapid and requires minimal or no sample pre-treatments. Such analytical tool can further be developed and adopted to a production line to allow optimal recovery of MDCM while maintaining permissible calcium or ash levels. 


Raman spectra were acquired on 79 meat-bone mixture samples from four different production days of industrial mechanical chicken deboning (Figure 1). Principal component analysis (PCA) of the spectra reveled qualitative variation between the samples collected from the four production days (Figure 2).  These variations were attributed, in addition to differences in calcium (or ash) content, differences in fatty acid compositions.  

Raman-based partial least squares (PLS) regression models were developed, based on the preprocessed spectra, for estimation of both ash and calcium content. The prediction model based on extended multiplicative scattering correction (EMSC) corrected Raman spectra afforded the lowest root mean square error of cross-validation and the highest coefficient of  

determination (Figure 3). 

PLSR correlation plot of calibration for determination of % calcium (A) and % ash (B) using the EMSC corrected Raman spectra.

Calsium and ash

The present work represents the first application of Raman spectroscopy for estimation of calcium and ash content in bone and meat mixtures from mechanical deboning of chicken. This technique holds a promising potential as industrially feasible on- or at-line tool for controlling quality of mechanically deboned chicken meat or similar food matrices. 


Further work in expanding the calibration data set as well as optimizing the data acquisition setup are required in order to develop a robust prediction models that can be used in an industrial process control. 

By: Jens Petter Wold    

EMSC-corrected Raman spectra of the 79 samples from mechanical deboning of chicken. All spectra are color-weighed according the % calcium.


Score plots from principal component analysis of the Raman data obtained for the 79 samples collected from four production days of mechanical chicken deboning.