Vestibular Compensation in Unilateral Patients Often Causes Both Gain and Time Constant Asymmetries in the VOR

Abstract

The vestibulo-ocular reflex (VOR) is essential in our daily life to stabilize retinal images during head movements. Balanced vestibular functionality secures optimal reflex performance which otherwise can be distorted by peripheral vestibular lesions. Luckily, vestibular compensation in different neuronal sites restores VOR function to some extent over time. Studying vestibular compensation gives insight into the possible mechanisms for plasticity in the brain. In this work, novel experimental analysis tools are employed to reevaluate the VOR characteristics following unilateral vestibular lesions and compensation. Our results suggest that following vestibular lesions, asymmetric performance of the VOR is not only limited to its gain. Vestibular compensation also causes asymmetric dynamics, i.e., different time constants for the VOR during leftward or rightward passive head rotation. Potential mechanisms for these experimental observations are provided using simulation studies.

Keywords: commissural neuron circuitry; plasticity; unilateral vestibular lesion; vestibular compensation; vestibulo-ocular reflex.

Figure 1

Eye velocity during 1/6 Hz head rotation, peak velocity 200 deg/s; (A) VOR of control subject (H₄), (B) VOR of subject diagnosed with right vestibular Ménière's disease (P₉). Note that head velocity is inverted for better alignment of velocity traces. Positive head velocity values refer to right side rotation and vice versa. Clinical analysis focuses on the envelope of eye velocity segments that follow the same pattern as the head velocity, here harmonic. Thus, the rapid pulses associated with fast phases are ignored. The results of data analysis related to these recording are presented in Figures 3, 4.

Figure 2

Assumed structure for the slow phase VOR: linear dynamics followed by a Hammerstein system (Jalaleddini and Kearney, 2013). H_v is head velocity (deg/s), H_c is the sensory signal (spikes/s) and E is conjugate eye position (deg). s refers to Laplace variable and T_v refers to the sensory time constant in the first block. G is the steady state gain and Bias is added to model the bias in VOR responses due to asymmetries in the second block. P refers to the pole of the central processing in the third block and the central processing time constant is: $T=\frac{-1}{P},P<0$.

Figure 3

(A) Estimated gain (g_co, g_ip) and (B) central processing time constant (T_co, T_ip) values for contra and ipsi-lesion rotations in 20 unilateral vestibular patients (P₁ − P₂₀). Ninety-five percent confidence intervals of the estimated values are shown with red bars which are computed based a Monte Carlo study from 200 repetitions of parameter estimates from randomly selected data intervals in the record—each sample set had the same number of data points.

Figure 4

(A) Estimated gain (g_l, g_r) and (B) central processing time constant (T_l, T_r) values for leftward and rightward rotations in 10 control subjects (H₁ − H₁₀). Ninety-five percent confidence intervals of the estimated values are shown with red bars which are computed based on 200 Monte-Carlo studies as described in Section 2.2.

Figure 5

T_index vs. G_index in unilateral patients and normal subjects. These indices refer to the relative difference of the estimated rightward-leftward parameters and emphasize the extent of asymmetries in the estimated VOR dynamics between the control and patient group.

Figure 6

Histogram of the estimated sensory time constant T_v related to the first block in Figure 2. (A) Patients; (B) Controls. There is no significant difference between the distribution of the estimated values in the two groups.

Figure 7

Bilateral model of horizontal slow phase AVOR in the dark adapted from Smith and Galiana (1991). The model includes sensory component modeled with high-pass dynamics followed by a static nonlinearity. Eye plant and PH are modeled with first order low-pass dynamics. PH projects efferent copies of eye position $E_{R,L}^{*}$ while C_R and C_L refer to the commissural projection weights and q refers to the sensory projection weight. More details are provided in the text.

Figure 8

Estimated dynamics of simulated VOR at acute stage of lesion with no compensation during ipsi- and contra-lesion rotations compared to intact VOR model. Left panel provides the gain, G and bias of the central VOR process in the brainstem. Note that in the intact model, the bias in the VOR response is zero and the ideal gain is −1, thus the red line is shifted up by 90 (spikes/s), the vestibular resting rate R_cR,cL, intentionally for comparison with the lesion case. The right panel shows the Bode plot of the expected form of VOR first order dynamic with time constant T = −1∕P, in the healthy condition.

Figure 9

Bilateral model of horizontal slow phase AVOR in the dark after unilateral lesion (right vestibular input) and compensation. In order to balance gain and remove bias after unilateral lesion, the resting activity of ipsi-lesion VN (R_2R) and commissural inhibition from the contra-lesion side (C_L) are increased (blue arrows). Moreover, commissural inhibition from the ipsi-lesion side (C_R) is decreased (red arrow). See Table 2 for parametric changes in the model.

Figure 10

Estimated dynamics of simulated VOR following compensation during ipsi- and contra-lesion rotations compared to intact VOR model. Left block reflects the gain, G and bias of the central VOR process in the brainstem and the right block shows the Bode plot of the estimated VOR first order dynamic with time constant T = −1∕P. The ipsi-lesion dynamics (blue) imbed a smaller time constant than during contra-lesion rotations. The latter can remain close to the control case.