The Fluorescence Fingerprint (FF)
The FF, also known as the Excitation Emission Matrix, EEM, is a set of fluorescence spectra which are acquired at consecutive wavelengths. With conventional fluorescence measurement, light of a certain wavelength is used as an excitation light and one spectrum is measured. With the FF, light in consecutive wavelengths are shone onto the sample and the fluorescence spectrum for each excitation light is measured. These fluorescence spectra are stacked 3-dimensionally. The FF shows a unique pattern for each constituent or material, just like a fingerprint.
Since the FF equals a whole set of fluorescence spectra, slight differences in fluorescence characteristics which cannot be determined from a single spectrum can be detected. It is also suited for measuring complex samples such as food, which are a mixture of many different fluorophors.
The measurement process is non-destructive and can be performed easily with a
fluorescence spectrophotometer.
Laser Scattering
Laser scattering measures the light scattering properties of target. Using light scattering properties, we can estimates physical properties of the target such as firmness and hardness. These physical properties are very important in foods, because we rely heavily on the "texture" of foods to evaluate their palatability. We want our apples to be "crisp", not "mealy"; we want breads to have their characteristic texture, whether it may be "soft as a pillow" or "crisp on the outside and chewy inside". Physical properties also affect how food materials can be processed, for example, processed cheese needs to be made from natural cheese materials that have a firm microstructure.
The measurement setup that we have constructed in our lab is based on spatially resolved diffuse reflectance, where the scattered and diffused laser light is captured with a camera. The system is relatively simple, and can be constructed without sophisticated equipment such as continuum lasers or hyperspectral cameras. Nevertheless, we have succeeded in estimating physical properties of fruits during storage with relatively high accuracy.
Hyperspectral imaging with the FF
In our laboratory, we are using the FF as the spectral data for hyperspectral imaging. Hyperspectral imaging is a technique that integrates conventional imaging and spectroscopy to attain both spatial information and spectral information from a sample (Gowen et al., 2007).
Fluorescence images of the sample are acquired with the imaging system shown on the right. There are two sets of band-pass filters which are used to control the excitation and emission wavelengths. Fluorescence images are acquired at all the combinations of excitation and emission wavelengths that construct the FF. The whole set of fluorescence images can also viewed as FFs of every pixel in the digital image. Since the FFs contain information on constituents existing at the corresponding points in the sample , mapping of specific constituents is possible by analyzing the FF data.
Visualization of multiple constituents in food
Here are some applications of FF imaging.
Gluten and Starch in bread dough
The palatability of bread is known to be influenced by its texture (the crunchiness of a baguette or silky softness of a brioche), and a large part of this texture relies on the way gluten and starch are distributed in the bread dough. The distributions of these constituents change rather quickly through the mixing process, so experienced artisans carefully monitor the mixing process to make the "optimum-mixed" dough.
Exactly how the distributions of gluten and dough change through the mixing process is a great interest for cereal chemists. From a more practical point of view, accurate determination of "optimum-mixing" through the visualization of these constituents would contribute greatly to the quality of bread product.
Gluten, starch and fat in pie pastry
Pie pastry was chosen because it contains protein (gluten), starch and fat, the three main constituents that create the structure of many foods. After the sample fluorescence images were acquired with the FF imaging system, the same sample was stained for protein (blue) and fat (orange). By comparing the FF image with the stained image, it was possible to validate that the imaging method was accurate.