Abstract
Characterizing the deformations undergone by the tongue during food oral processing could help to better understand how texture sensations are perceived. In this article, we propose to study the potential of ultrasound (US) imaging to monitor the deformations undergone by artificial tongues during compression and shear of agar food gels. Four polyvinyl alcohol cryogels were used as artificial tongues (two levels of roughness and two levels of stiffness), while three agar gels of different concentrations were considered as model foods.
Throughout the experiments, US images were acquired from a transducer array positioned underneath the artificial tongue, while force signals were obtained from a multi- axes sensor located above an artificial palate plate. Image analysis first consisted of tracing the contour of the dorsal surface of the artificial tongue. It was thus possible to observe how the deformations are distributed between the artificial tongues and the agar gels and to follow over time the heterogeneity of this distribution along the axis of the transducer array. Then, Particle Image Velocimetry (PIV) analysis was conducted to characterize the velocity fields related to deformations within the artificial tongue.
In particular, the horizontal component of the velocity was studied during the shear movements and allowed one to distinguish static and dynamic friction phases, and to highlight the deformation gradients in the bulk of the artificial tongue. Such US method can provide a better understanding of the impact of the mechanical properties of food gels on the stimulation of mechanoreceptors responsible for translating mechan
ical stimuli into sensory perceptions.
Download the full article as PDF here Ultrasound Imaging of Artificial Tongues During Compression and Shearing of Food Gels on a Biomimetic Testing Bench
2 | Materials and Methods
2.1 | Food Gels
Three different gels were considered as model foods in the study, differing in their concentration in agar (Ag), the main polymer used in their formulation. The preparation followed a previously described protocol (Srivastava, Mantelet, et al. 2021). In short, sucrose (Daddy, CristalCo SAS, Paris, France) was added and dissolved in ultrapure water at 10% g/l (w/w) and underwent stirring at room temperature (20°C) for 30 min. Then, edible agar (SAS Nature and Plants, Magescq, France) was added to the water and sucrose solution in concentrations of 0.45%, 0.60%, and 1.00% g/l (w/w) and heated at 95°C for 45 min under continuous magnetic stirring, with it being covered with aluminum foil. After this, it was poured into silicone molds (ELASTOMOULE, De Buyer, Le Val- d’Ajol, France) and kept in the refrigerator at ~5°C for a duration of 18 h. Two types of silicone molds were used: cylindrical molds were used to design samples suitable for mechanical characterizations, while cubical ones were consid ered to prepare samples dedicated to experiments on the biomimetic test bench. Cylindrical samples had a diameter of 25 mm and a height of 20 mm, while cubical ones had dimensions equal to 25 × 25 × 20 mm3 (width, depth, and height). Before the experiments, food gels were left at room temperature for 15 min and taken out from their silicone molds. After this, cylindrical gels went through a compression test using a texture analyzer (TA.XT plus, Stable Micro Systems, Surrey, United Kingdom). The samples were subjected to a deformation of 20% (relative to their initial height) at the speed of 10 mm/s. Stress vs. strain slope was characterized for the estimation of the Young’s modulus (between 2.5% and 5%, over which linear elasticity could be observed). Five replicates were performed for each of the three types of gels. The composition of the three different gels and the obtained values of Young’s modulus are summarized in Table 1. The gels are referred to according to their concentration in agar: Ag0.45, Ag0.60, and Ag1.00.
2.2 | Artificial Tongues
A total of four artificial tongues were manufactured and used in the study, varying both in terms of roughness and rigidity. The roughness of the artificial tongues was varied to account for the topography of the tongue surface, which is primarily induced by the filiform papillae that cover the dorsal surface. In previous work, roughness properties were found to strongly affect the mechanical interactions between the artificial tongue and
food, both during compression (Srivastava, Stieger, et al. 2021) and shear experiments (Glumac et al. 2023). Furthermore, this roughness greatly affects the transmission of ultrasound waves at the tongue- food interface, due to air bubbles trapped in between (Mantelet, Restagno, et al. 2020). It was therefore all the more important to take this factor into account in the present study.
The artificial tongues were Polyvinyl Alcohol (PVA) cryogels, prepared according to previously established protocols (Srivastava, Mantelet, et al. 2021; Srivastava, Stieger, et al. 2021). PVA powder (MW 89,000–98,000, 99% hydrolyzed, Sigma Aldrich, Saint- Louis, USA) was dissolved in ultrapure water at 10% (w/w) under magnetic stirring for 2 h at 80°C. After this, the temperature of the solution was cooled down to room temperature (20°C). In the present study, the main change made to the protocol consisted of adding 1% (w/w) cellulose particles with
a size distribution of 63–126 μm (Cellets 90, IPC GMBH & Co., Dresden, Germany) during the cooling step under continuous magnetic stirring. The scattering properties of these particles make it possible to increase the contrast of ultrasound images and to better observe the fields of deformation in the bulk of the artificial tongues. The obtained solution was poured into rectangular molds of 80 × 45 × 25 mm3 (width, depth, and height), the lower surface of which was covered with sandpaper to roughen the corresponding surface of the artificial tongue. Two references of sandpaper (Leman, Saint- Clair- de- la- Tour, France) were used (P100 and P36), leading to artificial tongues surfaces hereafter referred to as “smooth” or “rough”. Profilometry characterizations performed in previous works (Glumac et al. 2023) made it possible to establish that the grit size of the sandpaper (550 μm for P36, 160 μm for P100) turns out to correspond approximately to the average width of the asperities (RSm parameter) of the obtained artificial tongues. In terms of roughness average height (Ra parameter), artificial tongues prepared with the same two types of sandpaper (P100 and P36) resulted in average asperity heights of 25 and 140 μm, respectively (Glumac et al. 2023). These values were found to match with data collected in vivo on human tongues, ranging from 20 to 120 μm (Uemori et al. 2012).
The molds were kept at −20°C for 16 h and thawed at 20°C for 8 h. The rigidity of the artificial tongues was varied by adapting the number of cycles of freezing and thawing undergone by the samples: two cycles for the samples referred to as “compliant”, four cycles for those referred to as “hard”. Finally, the artificial tongues were unmolded and immersed in water using plastic containers, hermetically sealed, and kept at room temperature. The rigidity of the artificial tongues was characterized following the same protocol as for the food gels: 20% strain rate at the
speed of 10 mm/s. Young’s modulus was determined from the stress/strain slope between 2.5% and 5.0%. The average and standard deviations of the Young modulus after five replicates were 25.57 ± 0.76 kPa for compliant artificial tongues and 47.01 ± 0.80 kPa for hard artificial tongues. The four obtained artificial tongues are labeled as follows: TH/S, TH/R, TC/S, TC/R (H: hard, C: compliant; S: smooth; R: rough, respectively).
Source: Miodrag Glumac, Jean-Luc Gennisson, Vincent Mathieu, Ultrasound Imaging of Artificial Tongues During Compression and Shearing of Food Gels on a Biomimetic Testing Bench, This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2025 The Author(s). Journal of Texture Studies published by Wiley Periodicals LLC., Journal of Texture Studies, 2025; 56:e70030, https://doi.org/10.1111/jtxs.70030
