The vibrotactile belt
In contrast to other vibrotactile systems used to support rehabilitation programs, the current system is designed to be used permanently when standing or moving to improve balance. The current developed system is ambulant (battery life > 16 h of continuous use) using 12 tactors positioned in a belt worn around the waist (Fig. 1). The tactors are activated via a microprocessor based on the output of a 6DOF motion and tilt sensor also incorporated in the belt. The transfer function is fully programmable per tactor and accessible via Bluetooth.
Pilot studies
The system was evaluated and further improved and developed based on the outcome of several pilot experiments since 2002. By this way the type of tactor, the material and construction of the belt were optimised to obtain an effective stimulation all over the waist, which appeared to be a big challenge. In a pilot study we observed that feedback over the waist proved to be superior to feedback to the head. In these pilot studies we also evaluated various outcome parameters in laboratory conditions (during and after use of the belt) to quantify the impact of vibrotactile feedback (sway angle, sway area, expert judgement of a double-blind gait analysis by video recording, Nintendo WII games, etc.). Improvement of all these parameters could be shown in these pilot studies, although with substantial variance. However, it was shown that the improvement was only significant if the patients wore the belt. Balance improvement disappeared almost immediately after the belt was removed or switched off. As a consequence and taking notice of a recent review using systems to aid and support rehab for patients with vestibular deficits [30], we decided to evaluate the clinical relevance of the use of the belt under daily life ambulatory conditions, and asked the patient to score their balance and mobility on a scale from 0 to 10 before and after 2 months of daily use of the belt. Although we are aware that this approach was subjective, we aimed to quantify the absolute effect of the belt in daily life.
Patient selection
Patients were included after informed consent when they suffered from:
-
1.
Severe Bilateral Vestibular Loss, defined as:
-
2.
mean peak slow phase velocity of no more than 5°/s in bilateral bi-thermal caloric irrigations with water (30 and 44 degrees Celsius)
-
3.
gain of less than 15% on rotatory chair test (sinusoidal stimulus, 0.1 Hz, peak velocity 90 °/s)
-
4.
pathological head impulse test results for the horizontal and vertical canals.
-
5.
Severe imbalance with a fear to fall and/or actual falls, despite undergoing various type of VR therapies previously
-
6.
Quality of health score (QHS) as derived from the SF-36 < 40%.
-
7.
Self-reported overall Mobility and Balance (MBS) score below 5 (test range 0–10).
The questionnaires SF-36 and MBS were submitted to patients during an interview at the Division of Balance Disorders at Maastricht University Hospital. The Short-Form Health Survey (SF-36) is a validated short-form health questionnaire composed of 36 items grouped into 8 variables: “physical functioning”, “role physical,” “bodily pain”, “general health”, “vitality”, “social functioning”, “role emotional” and “mental health,” assessing both mental and physical health [31, 32]. For each variable, item scores were coded, summed, and transformed on a scale from 0 (worst possible health state measured by the questionnaire) to 100% (best possible health state). It is a generic health measure, rather than disease-specific and used here as a general measure expressed in a quality of health score (QHS). The validated Dutch version of the SF-36 was used in this study [33].
The MBS was a number given by the patient on a scale from 0 to 10, as a subjective grading of his or her overall mobility and balance in daily life and is considered as a more specific outcome measure in this study. The intra-individual change of the MBS was used as a first approximation of the impact of the balance belt.
Patients with neurological, psychiatric or orthopaedic disorders, reduced proprioceptive sensitivity, or impaired vision were excluded.
Study 1: placebo controlled evaluation of vibrotactile feedback using 5 different feedback stimulation patterns.
Five different feedback patterns were applied to investigate the most effective feedback and to allow a placebo-controlled evaluation of the impact of the balance belt. The patterns were defined as follows:
-
1.
As soon as the trunk inclination exceeded 2.5° relative to the gravity vector, the tactor towards which the trunk is inclined is activated: correct indication of the inclination direction. When the sensor returns within a tilt angle of 1.5°, the tactor will become silent again.
-
2.
Placebo: as soon as the trunk inclination exceeds 2.5°, in 1 out of 3 times the “correct” tactor towards which the trunk is inclined is activated. However, in 2 out of 3 times one out of the other 11 tactors is activated (at random): indication of the correct inclination direction in only 33.3% of the cases. When the sensor returns within a tilt angle of 1.5°, any activated tactor will become silent again.
-
3.
Placebo: as soon as the trunk inclination exceeds 2.5°, one out of the 12 tactors (at random) is activated: only feedback regarding the trunk inclination but no feedback of the inclination direction. When the sensor returns within a tilt angle of 1.5°, any activated tactor will become silent again.
-
4.
Placebo: tactors are activated at random, with at random intervals of 1–3 s, without any correlation to the trunk inclination: at random vibration but no feedback regarding inclination.
-
5.
Placebo: null mode: no vibration
We choose stimulus feedback pattern 1 for the tactor on the basis of the results of our pilot study: a 300 Hz sinusoidal signal delivered in sequences of 150 ms with a repetition rate of 4 Hz. Patients experience a vibratory pulse from a tactor when they tilt more than 2.5° towards that tactor. The stimulus is turned off when their tilt becomes less than 1.5° towards that sensor (Fig. 1).
A reset bottom allowed users to align the belt orientation with the gravity vector after mounting the device or when they needed to adjust their default body position (standing, sitting, car driving, etc.).
Based on the inclusion and exclusion criteria patients were selected and wore the belt for 1 day in the hospital using the standard feedback pattern 1. Five patients were selected that indicated that they experienced a clear improved balance and volunteered to participate in a double-blind “placebo-controlled” study using the belt programmed in 5 different feedback patterns, 1–5, defined as follows.
The patients wore the personalised belt with a specific feedback pattern for 6 weeks during all daily life activities. Each patient wore the same belt but with all 5 different feedback patterns for 6 weeks in random order. After each 6 weeks period, an interval of 2 weeks was taken without belt. The sequence of the 5 patterns was randomised over the 5 patients. The belts were programmed by a technician that annotated the codes of each belt. Patients, examiner and technician were blind regarding the specific feedback pattern used for a patient at any moment of the study. After 8 weeks the feedback pattern in the belt was changed according to the randomised protocol. At the start (t = 0), and after 2 (t = 2), after 4 (t = 4) and after 6 (t = 6) weeks the impact of the belt upon balance and gait was evaluated by the history, multiple tests and questionnaires.
The tests were:
-
Romberg eyes open eyes closed (measuring sway area and sway velocity)
-
Normal walking, with eyes open and eyes closed (video registration with blinded observation by 3 experts)
-
Heel to toe walking, with eyes open and eyes closed (video registration with blinded observation by 3 experts)
-
Ski-game on the Nintendo balance platform
-
QOL score, Dutch version of the Short Falls Efficacy Scale-International (2007), Dizziness Handicap Inventory
-
MBS score
Statistics was performed using SPSS, paired t tests.
Study 2
Based on these inclusion and exclusion criteria 39 patients were preselected. Subsequently, patients wore the belt for 2 h and were asked to indicate whether they experienced a clear benefit and wanted to participate in the 2-month study. Thirty-one patients, all indicating that they experienced a clear benefit, volunteered for participation in the study and signed an informed consent. Eight patients experienced no clear benefit of the belt during the 2 h try out and did not want to participate in the study.
After inclusion, all 31 patients were seen before and 1 month after the use of the belt in the outpatient clinic of the vestibular department by one of the primary investigators (Herman Kingma and Lilian Felipe). Before patients received the balance belt for a 1-month use, they had to indicate their mobility and balance score (MBS). They were asked to wear the belt continuously if they were standing or moving. A short instruction was given on how to use the belt. All patients started to use the belt without any problem and intuitively responded to the vibratory stimulus by adjusting their posture. After the trial, patients visited the outpatient department to return the belt. They were asked to indicate their MBS again. They were also actively asked to give additional comments about their subjective experiences with the belt.
Statistics
Statistics was performed using SPSS, paired t tests.
Ethical considerations
The procedures in this investigation were in accordance with the legislation and ethical standards on human experimentation in the Netherlands and in accordance with the Declaration of Helsinki (amended version 2013). Approval was obtained from the ethical committee of Maastricht University Medical Center. All procedures were performed at the Maastricht University Medical Center.