Real soft bites made by a model tongue to better assess food texture

Original paper: Compression Test of Food Gels on Artificial Tongue and Its Comparison with Human Test

Content review: Arthur Michaut
Style review: Heather S. C. Hamilton

Food texture, or how soft or hard your food is, is usually assessed with a machine that compresses food between two metallic plates. This instrument oversimplifies what really happens in our mouth, because the stiffness of the metallic parts does not mimic the natural deformability of the tongue. This leads to a strong deviation of the results of these experiments from sensory evaluations of food texture by humans. Practically, this can lead to hard-to-swallow food being considered soft and safe for people with dental or swallowing problems. A team of Japanese scientists tried to tackle that problem by developing a synthetic tongue which would lead to better food texture assessment with a test machine. But what are the important properties that characterize a successful model tongue?

To answer this question, the researchers compressed the tongues of seven brave human subjects using a clamp equipped with a force sensor like the schematic in Figure 1a. The human tongue is considered an elastic material, meaning that it can be deformed (by compression for example) and will return to its original size and shape after the force causing the deformation is removed. Deformation is often measured as percent strain, or the ratio of deformed size to original size of an object. Using this compression technique, which seems a bit medieval, and with the help of the human subjects, the researchers determined how the human tongue responds to an applied strain of up to 20%. Understanding the mechanical properties of the tongue means it can be modeled by an equivalent soft material and tested in a similar manner, as shown in Figure 1b. 

Figure 1: Schematic of the steps to design an in vitro system for tongue-palate compression tests. Tongue elasticity is first measured with a clamp (a). A silicone rubber tongue with a similar elasticity is then put on the bottom plate of a compression test machine to mimic the tongue (b). The metallic top plate will act as the palate, or the hard roof of the mouth.

The synthetic tongue produced for this study was a cylinder made of silicone rubber with an elasticity that is similar to the elasticity of the real tongues of seven human subjects. Remarkably, the tongues of seven subjects were considered a large enough sample size to determine the elasticity of an average human tongue, with resulting measurements close to values reported in the literature. In the food texture assessment device, the top metallic plate is left unaltered to mimic the palate. Pieces of agar gel, a jelly-like material made from seaweed, are used as a test food product. Several agar gel samples were prepared, each requiring some force to fracture, breaking into smaller pieces. Mechanical tests were performed with the in vitro tongue-palate system, while in vivo testing was determined by the same seven subjects who had their tongues clamped in the compression test. The in vivo results determined if the subjects were able to fracture the gel by tongue-palate compression or if they needed to chew the gel. These results were compared to the in vitro observations.  

The most interesting result highlighted by the study, shown in Figure 2, is the threshold above which mastication, the action of chewing with your teeth, was needed to break the agar gel into edible pieces. At a strain of 10%, if the agar gel sample was more deformed than the silicone rubber of the artificial tongue, the agar gel would fracture to the consistency of mashed potatoes. Otherwise, if the silicone rubber was more deformed than the agar gel sample, the gel sample would remain intact, meaning that a force higher than the one provided by tongue-palate compression would be needed for fracture, such as the force provided by mastication. Indeed, during tongue-palate compression, both the tongue and the model food are deformed under the action of reciprocal forces.

Figure 2: Strain profiles of the agar gel and the silicone rubber over time under compression by the tongue-palate model. At 10 % strain, if the agar gel curve is below the silicone rubber curve (a, top), which means that it is less deformed, then it will be more resistant to compression and will not collapse (a, bottom). If the opposite scenario takes place (b, top), then the gel will fail (b, bottom).

This is exactly how the human tongue acts like a mechanical sensor, thanks to the difference of stiffness between the organ’s tissues and the material probed. The resulting feeling will then influence the strategy used by our mouth to break down food in small ingestible (soft)bites. Food with a soft texture will only need this tongue-palate compression to be edible. Alternatively, foods with a tougher texture will need fragmentation through mastication. It is to be noted that the strategy used by the subjects eating the same agar gels matched the results of the in vitro compression tests: if the agar gel collapsed during the test, only tongue-palate compression was needed. 

Therefore, food products that would fracture during this compression test can be considered soft enough to be edible without chewing. Even if this system is not a perfect model since the synthetic tongue does not replicate the complex arrangements of different tissue layers present in the human tongue and a model saliva is not included, it can be of great help to design safe food for people with mastication problems. 

Leave a Reply