Objective analysis and prediction of texture perception of yoghurt by hybrid neuro-numerical methods

The trend for health-conscious nutrition leads to consumer interest in low-fat dairy products. For consumer acceptance low-fat yoghurts require sensorial characteristics similar to full-fat products. Estimation of texture attributes demands sensory investigations. The present contribution focuses on objective analysis and prediction of mouth feel attributes during swallowing of yoghurt. The applied hybrid neuro-numerical approach was based on numerical simulations, cognitive algorithms and experimental sensorial investigations. The swallowing of liquid foods leads to fluid mechanical interactions of food and tongue/palate inducing specific texture perception. By numerical simulations fluid mechanical interactions of yoghurt and mouth surface were determined and evaluated in a three-dimensional tongue–palate model. The numerical simulations revealed time dependent characteristic pressure and shear stress distributions at the surface of tongue and palate induced by the flow of yoghurt during swallowing. Prediction of sensorial texture attributes was successfully achieved by coupling results of numerical simulations with sensory evaluation using cognitive algorithms.