Evaluating the use of a balance prosthesis during balance perturbations in children and young adults with cochleovestibular dysfunction

Abstract

Study objectives were to: (1) quantify stability in children and young adults using cochlear implants with concurrent cochleovestibular dysfunction (CI-V) during balance perturbations and (2) to assess effects of an auditory head-referencing device (BalanCI) on their stability. The BalanCI provides auditory feedback via cochlear implants to cue posture and potentially avoid falling in children with CI-V. It was hypothesized that children and young adults with CI-V respond with larger movements to floor perturbations than typically-developing peers (controls) and that BalanCI use decreases these movements. Motion in response to treadmill perturbations was captured by markers on the head, torso, and feet in eight CI-V and 15 control participants. Stability (area under the curve of motion displacement) and peak displacement latencies were measured. The CI-V group demonstrated less stability and slower responses than the control group during medium and large backwards perturbations (p's < 0.01). In the CI-V group, BalanCI use improved stability during large backwards perturbations (p < 0.001), but worsened stability during large sideways perturbations (p's < 0.001). Children and young adults with CI-V move more to remain upright during perturbations than typically-developing peers. The BalanCI has potential to aid physical/vestibular therapy in children with CIs who have poor balance.

Figure 1

(a) Placement of head, upper body, pelvis, and foot markers. (b) Representation of markers in space as measured by motion capture software, and assessment of translational (anteroposterior and mediolateral) and rotational (pitch and roll) movements. (c) Treadmill set-up with harness. (d) Waveforms were created for translational and rotational motion for each marker cluster, and AUC was determined as a measure of overall magnitude and duration of movements. Pictured is an example of AUC during a large backwards perturbation of a pelvis marker in one participant in the CI-V group, and one participant in the control group.

Figure 2

(a) Overall performance on the BOT-2 by group on the x-axis, and age- and sex-scaled total BOT-2 scores on the y-axis. (b) Performance on BOT-2 static tasks with task number on the x-axis, and time to fall (s) on the y-axis. The static tasks are (1) standing in tandem stance on a stable surface with eyes open, (3) standing on one foot on a stable surface with eyes open, (4) standing in tandem stance on a stable surface with eyes closed, (6) standing on one foot on a stable surface with eyes closed, (7) standing on one foot on a balance beam with eyes open, (8) standing heel-to-toe on a balance beam with eyes open, and (9) standing on one foot on a balance beam with eyes closed. (c) Performance on the BOT-2 dynamic tasks with task number on the x-axis, and number of successful steps achieved on the y-axis. The dynamic tasks are (2) walking forward on a line, and (5) walking heel-to-toe on a line. The CI-V group (n = 8) is indicated in orange, and control group (n = 15) in blue. Error bars represent standard error of the mean, and points represent individual datapoints. Significance differences are indicated: **p < 0.01, ***p < 0.001.

Figure 3

(a) Translational displacement (cm) over time (s) and (b) rotational displacement (°) over time (s) collapsed over all motion capture clusters and BalanCI conditions for small, medium, and large treadmill perturbations in the forwards, backwards, leftward, and rightward directions. Anteroposterior axis and pitch plane are indicated for forwards and backwards perturbations, and mediolateral axis and roll plane are indicated for leftwards and rightwards perturbations. Translational motion: positive values are in the same direction as treadmill motion. Rotational motion: positive is pitch upwards and roll rightwards. (c) Absolute AUC for translational motion (cm*s) and (d) rotational motion (°*s) across small, medium, and large treadmill perturbations, plotted as mean treadmill velocity (cm/s) and collapsed across marker and BalanCI condition. The CI-V group (n = 8) is in orange, and control group (n = 15) in blue. Error bars represent standard error of the mean, and points represent means. Significance differences are indicated: **p < 0.01, ***p < 0.001.

Figure 4

(a) Translational displacement (cm) over time (s) and (b) rotational displacement (°) over time (s) collapsed over all motion capture clusters for small, medium and large treadmill perturbations in the forwards, backwards, leftward, and rightward directions in the CI-V (n = 8) and control (n = 15) groups, with BalanCI off and on. Anteroposterior axis and pitch plane are indicated for forwards and backwards perturbations, and mediolateral axis and roll plane are indicated for leftwards and rightwards perturbations. Translational motion: positive values are in the same direction as treadmill motion. Rotational motion: positive is pitch upwards and roll rightwards. (c) Absolute AUC for translational motion (cm*s) and (d) rotational motion (°*s) across small, medium, and large treadmill perturbations, plotted as mean treadmill velocity (cm/s) and collapsed across markers. BalanCI off is in red, and on in blue. Error bars represent standard error of the mean, and points represent means. Significance differences are indicated: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5

(a) Translational displacement (cm) over time (s) and (b) rotational displacement (°) over time (s) collapsed over all motion capture clusters and BalanCI conditions for small, medium, and large treadmill perturbations in the forwards, backwards, leftward, and rightward directions. Anteroposterior axis and pitch plane are indicated for forwards and backwards perturbations, and mediolateral axis and roll plane are indicated for leftwards and rightwards perturbations. Translational motion: positive values are in the same direction as treadmill motion. Rotational motion: positive is pitch upwards and roll rightwards. The shaded region represents the region of interest for peak selection. Peak 1 latency (s) for (c) translational and (d) rotational motion across small, medium, and large treadmill perturbations, plotted as mean treadmill velocity (cm/s). Peak 2 latency (s) for (e) translational and (f) rotational motion across perturbation degrees. Inter-peak latency (P2 latency – P1 latency) for (g) translational and (h) rotational motion across perturbation degrees. The CI-V group (n = 8) is in orange, and control group (n = 15) in blue. All measures are collapsed over motion capture clusters. Error bars represent standard error from the mean, and points represent means. Significance differences are indicated: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

(a) Translational displacement (cm) over time (s) and (b) rotational displacement (°) over time (s) collapsed over all motion capture clusters for small, medium and large treadmill perturbations in the forwards, backwards, left and right directions for the CI-V group (n = 8) and control group (n = 15) with the BalanCI off and on. Anteroposterior axis and pitch plane are indicated for forwards and backwards perturbations, and mediolateral axis and roll plane are indicated for leftwards and rightwards perturbations. Translational motion: positive values are in the same direction as treadmill motion. Rotational motion: positive is pitch upwards and roll rightwards. The shaded region represents the region of interest for peak selection. Peak 1 latency (s) for (c) translational and (d) rotational motion across small, medium, and large treadmill perturbations, plotted as mean treadmill velocity (cm/s). Peak 2 latency (s) for (e) translational and (f) rotational motion across perturbation degrees. Inter-peak latency (P2 latency – P1 latency) for (g) translational and (h) rotational motion across perturbation degrees. All measures are collapsed over motion capture clusters. BalanCI off is in red, and on in blue. Error bars represent standard error of the mean, and points represent means. Significance differences are indicated: *p < 0.05, **p < 0.01, ***p < 0.001.