Quantitative Visualization of the Distribution of Gluten and Starch in Model Dough with Fluorescence Fingerprint Imaging

The distribution of gluten and starch in wheat flour dough influences qualities of bread such as texture, and a method to visualize these distributions easily, accurately, and nondestructively would be of great value. The objective of this study was to develop an imaging method that visualizes the precise quantity of these two constituents, using the Fluorescence Fingerprint (FF). The FF is a set of fluorescence spectra acquired at consecutive excitation wavelengths, and its pattern contains abundant information on the constituent measured. Model dough samples were made with different ratios of gluten, starch, and water, and fluorescence images of the samples were acquired at the multiple combinations of excitation and emission wavelengths that construct the FF. A partial least squares regression (PLSR) model was built with the average FF of each sample and its gluten and starch ratios as the explanatory and objective variables, respectively. This model enabled the prediction of the gluten and starch ratios from the FF. The model was applied to the FFs of each pixel of the samples images and the gluten and starch ratios of the corresponding point in the sample was predicted. Finally, each pixel was coloured according to the predicted ratio, thereby creating images that quantitatively visualized the distribution of gluten and starch.