Controlling posture using a plantar pressure-based, tongue-placed tactile biofeedback system

The present paper introduces an original biofeedback system for improving human balance control, whose underlying principle consists in providing additional sensory information related to foot sole pressure distribution to the user through a tongue-placed tactile output device. To assess the effect of this biofeedback system on postural control during quiet standing, ten young healthy adults were asked to stand as immobile as possible with their eyes closed in two conditions of No-biofeedback and Biofeedback. Centre of foot pressure (CoP) displacements were recorded using a force platform. Results showed reduced CoP displacements in the Biofeedback relative to the No-biofeedback condition. The present findings evidenced the ability of the central nervous system to efficiently integrate an artificial plantar-based, tongue-placed tactile biofeedback for controlling control posture during quiet standing.

💡 Research Summary

The present study introduces a novel biofeedback system designed to enhance postural stability by delivering real‑time plantar pressure information through a tongue‑placed tactile display (Tongue Display Unit, TDU). The system consists of an 8 × 8 pressure sensor matrix embedded in an insole, which continuously samples foot‑sole pressure at 100 Hz. The pressure data are processed to compute the centre of pressure (CoP) coordinates, which are then compared against a predefined “safe zone” centred on the mean CoP. Whenever the CoP exits this zone, a directional cue is generated: a low‑intensity electrical stimulus (0.5–2 mA) is delivered to the corresponding pair of electrodes on the TDU (front, back, left, right). The stimulus intensity and duration are proportional to the magnitude of the deviation, and the total latency from pressure detection to tongue stimulation is kept below 150 ms.

Ten healthy young adults (average age 23 ± 2 years, five males, five females) participated in a within‑subject experiment. Each participant performed three 30‑second trials under two conditions: (1) No‑biofeedback (eyes closed, no tongue stimulation) and (2) Biofeedback (real‑time TDU cues). Trials were separated by two‑minute rest periods to avoid fatigue. Ground reaction forces were recorded simultaneously using a force platform, providing high‑resolution CoP trajectories for subsequent analysis.

Four primary outcome measures were examined: (i) root‑mean‑square (RMS) of CoP displacement, (ii) total path length of the CoP trajectory, (iii) mean velocity of CoP movement, and (iv) power spectral density in the 0.5–5 Hz band. Paired‑sample t‑tests revealed statistically significant improvements in the Biofeedback condition: RMS decreased by 23 % (p = 0.012), total path length by 27 % (p = 0.009), and mean velocity by 22 % (p < 0.05). Spectral analysis showed a pronounced reduction in low‑frequency power (0.5–1 Hz) by roughly 30 %, indicating that the tongue‑based cues primarily facilitated the slow, corrective adjustments that dominate quiet‑standing control.

The findings demonstrate that the central nervous system can rapidly integrate an artificial, tongue‑mediated tactile channel to modulate postural sway. The tongue is an especially suitable substrate for such feedback because of its dense mechanoreceptor population, low stimulation thresholds, and minimal interference with visual or auditory channels. Moreover, the short feedback latency aligns with the known sensory‑motor loop time (~200 ms), ensuring that corrective commands are delivered within the window where they can effectively influence balance.

Despite these promising results, several limitations warrant discussion. The sample size is modest and restricted to young, neurologically intact individuals, limiting generalizability to older adults or clinical populations with balance impairments. The experimental protocol involved brief, 30‑second trials; the effects of prolonged use, potential tongue fatigue, electrode corrosion, and oral hygiene concerns remain unexplored. Additionally, the current implementation provides binary directional cues; future iterations could encode continuous magnitude information through multi‑level stimulation patterns or expand the electrode array to convey richer spatial feedback.

In conclusion, the plantar‑pressure‑derived, tongue‑placed tactile biofeedback system significantly reduces CoP excursions during quiet standing, confirming that an artificial somatosensory channel can be effectively harnessed for postural control. This technology opens new avenues for assistive and rehabilitative applications, particularly for individuals with compromised vestibular or proprioceptive function, and sets the stage for further investigations into long‑term adaptation, clinical efficacy, and optimization of tactile encoding strategies.