1 Department of Applied Mathematics and Computer Science, Technical University of Denmark2 Image Analysis & Computer Graphics, Department of Applied Mathematics and Computer Science, Technical University of Denmark3 Statistics and Data Analysis, Department of Applied Mathematics and Computer Science, Technical University of Denmark4 University of Copenhagen
Using novel X-ray techniques, based on grating-interferometry, new imaging modalities can be obtained simultaneously with absorption computed tomography (CT). These modalities, called phase contrast and dark field imaging, measure the electron density and the diffusion length of the sample. Enhanced contrast capabilities of this X-ray technique makes studies on materials with similar attenuation properties possible. In this paper the focus is set on processing grating-based X-ray tomograms of meat emulsions to quantitatively measure micro-structural changes due to heat treatment. The emulsion samples were imaged both in a raw and cooked state. Additionally, different fat types were used in the emulsions in order to compare micro-structural differences when either pork fat or sunflower oil was used. From the reconstructed tomograms the different ingredients in the emulsions were segmented using a multivariate segmentation method. From this, a quantitative analysis was performed between the different samples, determining properties such as percentage object volumes and cooking loss. Additionally, the porosity, degree of anisotropy and average structure thickness of the protein networks were determined. Analyzing the multivariate dataset instead of the single univariate absorption modality gave superior segmentation results. The quantitative analysis of the micro-structure gives insight to how both heat treatment, and the use of different lipid types, affect the final protein network.