Why Calibration Matters in Rotatory Chair Testing

Rotatory chair testing measures the vestibulo-ocular reflex (VOR) by rotating a subject and recording the resulting eye movements. Because the clinical value of the test depends on accurate conversion of recorded eye signals into degrees of eye rotation, and of turntable motion into degrees per second of head rotation, every result is only as trustworthy as the calibration that produced it. A rotatory chair calibration error can shift VOR gain, alter computed phase, distort asymmetry between rightward and leftward responses, and change the apparent time constant derived from velocity step testing.

Calibration errors are rarely dramatic in isolation. A few percent of gain error, a small timing offset, or a slow drift in an eye-tracking channel may look unremarkable on a single record while systematically biasing results across a clinic. Recognising the sources of these errors, and building checks into the testing routine, is therefore as important as the interpretation of the numbers themselves.

The Three Calibration Domains

In a typical rotary chair laboratory, calibration operates in three separate domains, each of which can introduce its own error.

• Oculomotor calibration: the relationship between the recorded eye signal and true eye position in degrees. This is usually established by asking the subject to fixate targets at known eccentricities while the recording system measures the resulting voltage or pixel displacement.
• Chair or turntable calibration: the relationship between the commanded rotation and the actual angular motion delivered. This depends on the tachometer, encoder, drive servo, and mechanical coupling of the chair.
• Temporal calibration: the alignment of eye and chair data in time. Sampling clocks, filter delays, and buffering can offset one channel relative to the other, which is particularly damaging for phase measurements.

Common Sources of Rotatory Chair Calibration Error

Eye Movement Recording Errors

• Non-linear gain of the eye channel, so that a calibration valid at 10 degrees is not valid at 30 degrees.
• Goggle slippage or electrode migration during the session, producing a step change in the baseline.
• Baseline drift in electro-oculography, or slow drift in video-oculography tracking when the pupil edge is partially obscured by the eyelid.
• Incorrect scaling when calibration is performed with a target at an assumed distance rather than a measured angle.
• Blink artefacts and partial lid closure being averaged into the calibration factor.
• Accommodation or convergence during fixation, which changes the effective eye position being measured.

Chair and Motion Errors

• Tachometer gain drift with temperature, or after service and re-mounting of the sensor.
• Encoder resolution limits that round small velocities to zero or produce quantisation steps.
• Belt slip, backlash, or compliant coupling between the motor and the chair platform.
• Servo lag that makes the actual velocity profile differ from the commanded profile, especially at higher frequencies.
• Incorrect assumption that the head and the chair move identically when head restraint is loose.

Timing and Signal Processing Errors

• Uncorrected delay introduced by anti-aliasing filters or smoothing algorithms.
• Mismatch between the sampling rate of the eye channel and the chair channel.
• Software assumptions about the start of a stimulus cycle that do not match the hardware trigger.

Consequences for Clinical Interpretation

A rotatory chair calibration error propagates directly into the reported metrics.

• VOR gain: an error in eye-channel gain or chair velocity scaling appears as a proportional error in gain. A subject with normal function may appear to have a mild bilateral vestibular weakness, or a genuine deficit may be masked.
• Phase: a timing offset of even a few tens of milliseconds produces a measurable phase lead or lag, mimicking the pattern seen in central vestibular dysfunction.
• Asymmetry: non-linear calibration or directional differences in goggle fit create apparent right-left asymmetries that are artefacts of the recording rather than signs of unilateral peripheral loss.
• Time constant: in velocity step testing, calibration drift during the slow-phase response distorts the decay curve and biases the estimated time constant.
• Test-retest reliability: inconsistent calibration between visits undermines the ability to track change over time, which is often the main reason for performing rotational testing.

Detecting Calibration Error

Several practical checks help identify a rotatory chair calibration error before it influences a clinical conclusion.

1. Perform oculomotor calibration before and after the rotational runs and compare the factors. A large change suggests drift or slippage during the session.
2. Verify calibration at more than one eccentricity to test linearity, not just a single central gain value.
3. Use a known external reference, such as a fixed visual target at a measured angle, to audit the eye channel independently of the software.
4. Check the chair tachometer against an independent measurement of angular displacement or velocity during routine servicing.
5. Inspect raw traces for baseline steps, clipping, or sudden offsets that would indicate a hardware or interface problem.
6. Repeat a standard stimulus and confirm that gain and phase are reproducible within the laboratory's expected tolerance.
7. Record calibration values in the report or the technical log so that outliers can be traced retrospectively.

Reducing Calibration Error

• Allow equipment to reach thermal equilibrium before testing, and follow the manufacturer's warm-up procedure.
• Fit goggles or electrodes carefully, secure them, and recheck position after the head restraint is tightened.
• Minimise head movement relative to the chair with an appropriately sized restraint and clear instructions to the subject.
• Discard or repeat runs in which blink artefacts, drowsiness, or excessive movement contaminate the record.
• Maintain a scheduled service programme for the turntable, encoder, tachometer, and eye-tracking hardware.
• Document software version, filter settings, and sampling rates, and revalidate the calibration chain after any update.
• Use consistent, written protocols so that calibration technique does not vary between operators.

Reporting and Interpretation

When calibration is uncertain, the safest approach is to state that uncertainty explicitly. Reporting gain and phase without the corresponding calibration values makes it difficult for a reader to judge how much confidence to place in a borderline result. Where a run is technically compromised, it is usually better to repeat the test than to interpret it, since a small rotatory chair calibration error can move a value from the normal range into the abnormal range, or the reverse.

Rotational testing remains a valuable tool for assessing the vestibulo-ocular reflex across a range of frequencies, for evaluating compensation after unilateral loss, and for distinguishing peripheral from central patterns. Its strength depends on disciplined calibration and on routine verification that the chair, the eye-tracking system, and the timing of the two are behaving as the software assumes.

Summary

Rotatory chair calibration error arises from three interacting domains: the eye movement recording chain, the turntable and its sensors, and the timing that links them. Errors in any of these domains distort VOR gain, phase, asymmetry, and time constant measurements. Regular verification, careful subject setup, scheduled maintenance, and transparent reporting of calibration values together limit the impact of these errors and preserve the clinical usefulness of rotary chair testing.