Anticipating food structure of meat products from mastication physics applying machine learning.

Alternatives to animal-based products are becoming more relevant. Most of those products rely at some stage on a structuring process; hence researchers are developing techniques to measure the goodness of the structured material. Conventionally, a typical sensory study or texture analysis by measuring deformation forces would be applied to test the produced material for its texture. However, meat alternatives and meat differ in more points than just the texture, making it hard to extract the isolated texture impression. To objectively obtain qualitative and quantitative differences between different food structures, evaluation of oral processing features is an upcoming technology which qualifies as promising addon to existing technologies. The kinematic data of the jaw and exerted forces regarding muscle activities are recorded during mastication. Resulting datasets are high in dimensionality, covering thousands of individual chews described by often more than ten features. Evaluating such a dataset could benefit from applying computational evaluation strategies designed for large datasets, such as machine learning and neural networks. The aim of this work was to assess the performance of machine learning algorithms such as Support Vector Machines and Artificial Neural Networks or ensemble learning algorithms like Extra Trees Classifier or Extreme Gradient Boosting. We evaluated different pre-processing techniques and various machine algorithms for learning models with regard to their performance measured with established benchmark values (Accuracy, Area under Receiver-Operating Curve score, F1 score, precision-recall Curve, Matthews Correlation Coefficient (MCC)). Results show remarkable performance of classification of each single chew between isotropic and anisotropic material (MCC up to 0.966). According to the feature importance, the lateral jaw movement was the most important feature for classification; however, all features were necessary for an optimal learning process.

Oral processing, rheology, and mechanical response: Relations in a two-phase food model with anisotropic compounds.

Food-material poses a challenging matrix for objective material scientific description that matches the consumers' perception. With eyes on the emerging structured food materials from alternative protein sources, objectively describing perceived texture characteristics became a topic of interest to the food industry. This work made use of the well-known methodologies of jaw tracking and electromyography from the field of "food oral processing" and compared outcomes with mechanical responses to the deformation of model food systems to meat alternatives. To enable transferability to meat alternative products, an anisotropic structuring ingredient for alternative products, high-moisture texturized vegetable protein (HM-TVP), was embedded in an isotropic hydrocolloid gel. Data of the jaw movement and muscle activities exerted during mastication were modeled in a linear mixed model and set in relation to characteristic values obtained from small- and large-strain deformation. For improvement of the model fit, this work makes use of two new data-processing strategies in the field of oral processing: (i) Muscle activity data were set in relation to true forces and (ii) measured data were standardized and subjected to dimensional reduction. Based on that, model terms showed decreased p-values on various oral processing features. As a key outcome, it could be shown that an anisotropic structured phase induces more lateral jaw movement than isotropic samples, as was shown in meat model systems.

Fibrous food and particle size influence electromyography and the kinematics of oral processing.

Structure-sensory relationships are essential for understanding food perception. Food microstructure impacts how a food is comminuted and processed by the human masticatory system. This study investigated the impact of anisotropic structures, explicitly the structure of meat fibers, on the dynamic process of mastication. For a general understanding of texture-structure relationships, the three typically used deformation-tests: Kramer shear cell-, Guillotine cutting- and texture-profile-analyses were conducted. 3D jaw movements and muscle activities of the masseter muscle were additionally tracked and visualized using a mathematical model. Particle size had a significant effect on jaw movements and muscle activities for both the homogeneous (isotropic) and fibrous (anisotropic) meat-based samples with the same composition. Mastication was described using jaw movement and muscle activity parameters determined for each individual chew. The adjusted effect of fiber length was extracted from the data, suggesting that longer fibers induce a more strenuous chewing in which the jaw undergoes faster and wider movements requiring more muscle activity. To the authors' knowledge, this paper presents a novel data analysis approach for identifying oral processing behavior differences. This is an advancement on previous studies because a holistic overview of the entire mastication process can be visualized.

Insights into characterizing and producing anisotropic food structures.

Besides the flavor profile of food, texture plays a major role in terms of the acceptance and likeability of food products. In contrast to gel-like homogenous isotropic structures, where the characterization is established and structure-texture mechanisms are well understood, there is still a lack of knowledge in the field of anisotropic complex food matrices. Food systems that show anisotropic properties in terms of macroscopic mechanical anisotropy as in grown meat, or mixed complex systems where anisotropic shaped particles or fibers are embedded into an isotropic matrix are challenging to characterize, hence the structure-texture correlation is not trivial to understand. In this paper, we bring together the state of the art of different anisotropic structures as a source of food, their formation in terms of structured plant proteins, and consequently the structure-texture correlation of those. Characteristic terms and properties to differentiate between anisotropic systems are introduced with the purpose to facilitate characterization of those. Based on the here provided terms and characteristics, further studies toward understanding such systems and their perception can be conducted. Beyond that, a first opinion on crucial influencing factors on the perception of anisotropic systems and their mechanistic background is provided.