Multispectral imaging techniques for evaluating physicochemical-, antioxidant, and sensory properties of Alaria esculenta and Saccharina latissima

Abstract

With the increased seaweed production worldwide, there is a need to focus on improved production practices to produce high-quality seaweed biomass, especially if the biomass is intended for high-quality products and human consumption. Multispectral imaging (MSI) is a novel technique used as a quality control tool within the food industry due to its rapid, real-time measurements. Therefore, the study aimed to assess the possibilities of using MSI as a quality control tool within seaweed cultivation to predict the physicochemical (including proximate composition, trace minerals, pH, and color), microbiological (total viable counts (TVC)), sensory, and antioxidant properties of A. esculenta and S. latissima throughout diverse processing and handling. The results showed differences in spectra between species, and species classification got 100 % accuracy when using a Support Vector Machine (SVM) with the spectral data. Furthermore, results indicated that the partial least square regression (PLSR) models developed with cross-validation of the MSI data effectively predicted multiple quality parameters, including pH ($R_{\mathrm{CV}}^2$ = 0.94, RMSECV = 0.278), carbohydrate content ($R_{\mathrm{CV}}^2$ = 0.89, RMSECV = 0.76), protein content ($R_{\mathrm{CV}}^2$ = 0.94, RMSECV = 0.12), ash content ($R_{\mathrm{CV}}^2$ = 0.80, RMSECV = 0.51), and phenolic content ($R_{\mathrm{CV}}^2$ = 0.992, RMSECV = 0.24). In addition, the results showed possibilities of using the technology to assess several sensory properties. In conclusion, the results show the potential of using the MSI technology as an effective quality control tool within the seaweed industry to simultaneously evaluate multiple physicochemical properties of brown seaweed biomass.