Application of vibrotactile feedback of body motion to improve rehabilitation in individuals with imbalance

Abstract

Background and purpose: Balance rehabilitation and vestibular or balance prostheses are both emerging fields that have a potential for synergistic interaction. This article reviews vibrotactile prosthetic devices that have been developed to date and ongoing work related to the application of vibrotactile feedback for enhanced postural control. A vibrotactile feedback device developed in the author's laboratory is described.

Methods: Twelve subjects with vestibular hypofunction were tested on a platform that moved randomly in a plane, while receiving vibrotactile feedback in the anteroposterior direction. The feedback allowed subjects to significantly decrease their anteroposterior body tilt but did not change mediolateral tilt. A tandem walking task performed by subjects with vestibulopathies demonstrated a reduction in their mediolateral sway due to vibrotactile feedback of mediolateral body tilt, after controlling for the effects of task learning. Published findings from 2 additional experiments conducted in the laboratories of collaborating physical therapists are summarized.

Results: The Dynamic Gait Index scores in community-dwelling elderly individuals who were prone to falls were significantly improved with the use of mediolateral body tilt feedback.

Discussion and conclusions: Although more work is needed, these results suggest that vibrotactile tilt feedback of subjects' body motion can be used effectively by physical therapists for balance rehabilitation. A preliminary description of the third-generation device that has been reduced from a vest format to a belt format is described to demonstrate the progressive evolution from research to clinical application.

Figures

Figure 1

Balance Vest or 2nd generation research test-bed. The device being worn by the female subject is autonomous and consists of three major parts. The wide white elastic band which rings the torso contains an array of 48 tactile vibrators (tactors). This array is shown at the left bottom. Direction is displayed by selection one of the 16 columns of tactors, while magnitude is displayed by selection one of three rows in that column. A 6 degree of freedom motion sensor package is mounted at the small of the back on top of the white band. The signals from the motion sensor unit are processed by a PC 104 computer that activates individual amplifiers connected to each of the 48 tactors. These electronic components and their battery power are mounted in two black leather holsters worn around the waist. Signals to and from the on-board computer are transmitted via WiFi to the laptop computer shown in the foreground. The experimenter (male subject) is thus able to monitor and to control the on-board computer remotely. The details are published in Wall and Weinberg.

Figure 2

Root mean square ML tilt during 24 tandem walking trails with and without vibrotactile tilt feedback, for 10 vestibulopathic subjects. Blue diamonds show control (no feedback) and green squares show trials using feedback. Three neighboring trials were averaged to yield 8 points for all 24 trials. Error bars show one standard deviation. Linear trends in the data are shown by dashed blue line for the control condition, while the solid green line shows the tilt feedback trend.

Figure 3

A) Example of M/L trunk tilt estimate time series for one DGI subtest (Walk while making vertical head movements) with feedback OFF and ON. B) Left and Center -- photos of a subject at the end of a DGI subtest (i.e. walk while making horizontal head movements). Note body alignment and position of the hands. C) Right -- change in average DGI score with feedback OFF (dashed red lines) and feedback ON (solid green lines) plotted versus the chance of falling in community based elders versus Dynamic Gait Index, scores from a model based upon Shumway-Cook et al.

Figure 4

Balance Belt or 3rd generation prosthesis prototype. Decreasing the number of tactile vibrators from 48 to four has resulted in a smaller, light-weight, belt-like device. The motion sensors, tactors, and supporting electronics are now contained within a flexible, moisture-resistant “skin” than can be worn under a blouse or shirt without being noticed. Four black plastic boxes house the electronic parts. Because the whole belt is stretchable, two sizes will fit the 99th percentile of the adult human population.