Development and validation of a biomimetic mastication simulator for dynamic texture evaluation for cooked rice

A biomimetic mastication simulator (BMS) was developed and integrated with a texture analyzer to reproduce key physiological features of mastication, including cyclic dental compression, saliva lubrication, and enzymatic hydrolysis. Chewing parameters were calibrated from in vivo measurements (38 cycles; 1.84 mL saliva containing 809.7 U·mL^-1α-amylase and 1038.69 μg·mL^-1 mucin). The BMS reproduced population-level bolus properties, achieving similarity indices of 71.43% for relative and 82.14% for cumulative particle-size distributions, and comparable glucose release between in vivo and in vitro boluses (14.91 vs 12.33 mmol/L). SEM and CLSM further confirmed comparable matrix erosion and starch-protein reorganization. Instrumental dynamics aligned with oral perception, with BMS-derived hardness and adhesiveness correlating strongly with sensory time-intensity profiles (R² = 0.8617 and 0.9258, respectively). Differences in oral breakdown and dynamic texture evolution of Indica (IR), Japonica (JR), and waxy (WR) rices were investigated by the BMS. IR, with high amylose content (26.8%), maintained a compact bolus structure, exhibited the highest dynamic hardness and chewing work, fragmented more slowly, and showed limited early hydrolysis. In contrast, WR, dominated by amylopectin (69.85%) and nearly amylose-free, underwent rapid structural collapse into fine, highly porous particles, exhibited the highest adhesiveness, and released approximately 2.5-fold more glucose than IR and JR. Collectively, these results demonstrated that varietal starch constituents governed the bolus structural disintegration and starch hydrolysis during oral processing, thereby shaping dynamic texture perception. The BMS provided a quantitative framework for structure-guided design of rice and starch-based foods with targeted sensory properties and early digestive outcomes.