Measurement of water fractions in freeze-dried shiitake mushroom by means of multispectral imaging (MSI) and low-field nuclear magnetic resonance (LF-NMR)

Shoaib Younas,Yu Mao,Changhong Liu,Mian Anjum Murtaza,Zeshan Ali,Liyang Wei,Wei Liu,Lei Zheng
DOI: https://doi.org/10.1016/j.jfca.2020.103694
IF: 4.3
2021-03-01
Journal of Food Composition and Analysis
Abstract:<p>The production of high-quality processed foods requires an in-depth understanding of water distribution in cellular spaces. This study validated the non-destructive investigation of water fractions in freeze-dried shiitake mushrooms using multispectral imaging (MSI) at 405−970 nm and low-field nuclear magnetic resonance (LF-NMR). The quantitative modeling of water fractions including partial least squares (PLS), back propagation neural network (BPNN) and least squares-support vector machines (LS-SVM), were employed. The LF-NMR analysis has shown the fast and easy removal of free water owing to weak bonding with tissue structure. Freeze drying left a maximum proportion of bound water and small amount of immobilized water at the end of process. Among the models, LS-SVM demonstrated the highest coefficient of prediction determination (<em>R<sup>2</sup></em><sub>p</sub>) in the quantitative estimation of water contents and was considered the best model obtaining <em>R<sup>2</sup></em><sub>p</sub> of 0.85, 0.90, 0.88, and 0.87 plus the lowest root mean square error of prediction of 12.61 %, 7.10 %, 18.13 %, and 4.55 % for free, immobilized, bound, and total water contents, respectively. MSI combined with chemometrics would be a promising technique for the non-destructive and fast monitoring of the water fractions. The findings of this study would be useful to retain the nutritional and quality attributes of agricultural commodities.</p>
food science & technology,chemistry, applied
What problem does this paper attempt to address?
The paper primarily explores the issue of non-destructive detection of moisture content in freeze-dried shiitake mushrooms using Multispectral Imaging (MSI) and Low-Field Nuclear Magnetic Resonance (LF-NMR) technology. Specifically, the research team observed the changes in moisture at different stages by freeze-drying shiitake mushrooms and conducted quantitative analysis of three different types of water: free water, immobilized water, and bound water using MSI and LF-NMR methods. The study employed three modeling methods—Partial Least Squares (PLS), Back Propagation Neural Network (BPNN), and Least Squares Support Vector Machine (LS-SVM)—to predict moisture content and compared the effectiveness of these models. The research results showed that during the freeze-drying process, free water was completely evaporated within 24 hours, while immobilized water significantly decreased after 24 hours. Ultimately, only about 0.22% of bound water remained in the shiitake mushroom samples. Among the models compared, the LS-SVM model performed the best, with prediction determination coefficients (R2p) of 0.85 for free water, 0.90 for immobilized water, 0.88 for bound water, and 0.87 for total moisture, along with the lowest Root Mean Square Error of Prediction (RMSEP). This indicates that MSI combined with chemometrics can serve as a rapid and non-destructive technique for monitoring moisture distribution in food, helping to maintain the nutritional value and quality characteristics of agricultural products.