Decoding the dynamic formation of bolus: A multimodal fusion and simulation framework for texture...

This study systematically analyzed the dynamic mechanisms through which food texture regulates bolus formation by integrating food and bolus properties, dynamic sensory perception, and multimodal physiological monitoring. A fermented biscuit model with a distinct hardness gradient (10.24-202.82 N) was established. Time-intensity and temporal dominance of sensations analyses revealed that the perception of soft biscuits began with "soft," while that of hard biscuits started with "crumbly/hard." All samples progressed through a "sticky/pasty" dominant phase in later stages, eventually tending toward a "spongy" sensation. Bolus analysis indicated that soft biscuits relied on a high-porosity structure for rapid saliva penetration but exhibited weaker cohesion, whereas hard biscuits formed a dense bolus through continuous mechanical breakdown and efficient binding with saliva. As hardness increased, biting force rose, salivary pH and mucin concentrations increased, peripheral physiological signals suggested a higher autonomic nervous load, and electroencephalography showed enhanced activation in sensorimotor and prefrontal networks. The innovatively proposed Multimodal Masticatory Dynamic Fusion and Analysis Model (MDFAM) enabled automatic three-stage division of the chewing process, revealing that increasing hardness prolonged the initial breakdown phase (S1) from 2.9 s to 6.8 s and shortened the pre-swallowing phase (S3) from 7.0 s to 1.6 s. Furthermore, a coupled dynamic simulation system was developed, which used decoded physiological signals to drive three-dimensional physical simulations, achieving for the first time the dynamic visualization of the entire process from bolus dispersion and mixing to aggregation. This completed an end-to-end modeling pipeline from multimodal perception to physical simulation.