vHIT Atlas

The setup

A vHIT test takes about ten minutes once you've done a hundred of them. The patient sits upright facing a wall target one metre away. You strap a lightweight head-mounted goggle on firmly — tight enough that it cannot slip when you give the head a sudden thrust — and run a brief calibration where the patient tracks dots on the wall with their eyes only. The software works out where the patient is looking. Then you stand behind the patient, hold their head with both hands, and deliver small, fast, unpredictable head thrusts in the planes of the canals you want to test.

That description sounds simple and it is — but every word in it matters. Goggles that slip, thrusts that are too slow or too predictable, head positions that miss the canal plane, patients who tense their neck and resist — every one of these introduces artefact. Most learning curves for vHIT plateau around 100 traces, and pitfalls cluster in identifiable categories^{[Hülse R 2020]}.

Delivering the thrust

A good vHIT head thrust is:

Small — 10 to 20° of total head displacement, no more.
Fast — peak velocity 150 to 250°/s. Below 100°/s the test is uninterpretable; above 300°/s the head signal saturates and clinicians often get hurt fingers.
Unpredictable — the patient must not know which way the next thrust is coming. Smooth, rhythmic thrusts let the smooth pursuit and pre-programmed saccade systems take over, and the trace is no longer a pure VOR measurement.
Aligned with the canal plane — yaw for lateral canals (head pitched down ~30° to bring the lateral canal into the horizontal plane), oblique forward/back for the vertical canals.

Calibration matters more than people think

The calibration step calculates where the pupil sits when the patient looks at each calibration target. Once that mapping is wrong, every subsequent eye-velocity measurement is also wrong. Common calibration failures: patient moves their head during calibration; the goggle slips between calibration and the first thrust; ambient light reflects off the cornea and the system mistracks the pupil.

If gain values look strange — particularly if you see asymmetry between rightward and leftward thrusts in the same canal — recalibrate before you start interpreting. Saving a bad calibration and analysing the resulting gains is the most common source of false-positive vHIT in the literature^{[Hülse R 2020]}.

Goggle slippage

The single most clinically important artefact is goggle slippage. When the strap loosens or shifts during a head thrust, the goggle moves relative to the head. The accelerometer in the goggle reads this slip as additional "head" movement, which the software then divides out of the eye velocity. The resulting gain is artefactually too high or too low depending on the direction of slip^{[Suh MW 2017]}.

Telltale signs of slippage: gain values that vary wildly between consecutive thrusts in the same direction; gain greater than 1.2 in any canal (true VOR gain almost never exceeds 1.1); a sudden change in gain partway through a recording session. If you see any of these, stop, re-fit the goggle, re-calibrate, and re-test. Trust no asymmetry that you have not been able to reproduce after refitting.

Vertical-canal testing

Vertical canals need head positions that the patient cannot easily anticipate. Two acceptable techniques: rotate the patient's head 45° to one side and deliver pitch thrusts (forward for one anterior canal, backward for one posterior); or keep the head facing forward and deliver oblique thrusts down-and-out (anterior) or up-and-in (posterior). The first technique is easier to learn but limits the patient's comfort; the second requires more practice but generates better data once mastered^{[MacDougall HG 2013]}.

Key teaching points

Goggle fit is everything. If it can slip, the data is suspect.
Thrust: small, fast (150–250°/s), unpredictable, in the canal plane.
True VOR gain rarely exceeds 1.1. Values above 1.2 are nearly always artefact.
Always look at individual traces, not just averaged gains.

Technique

The setup

Delivering the thrust

Calibration matters more than people think

Goggle slippage

Vertical-canal testing

The seven common artefacts

When to refer back to calorics

Key teaching points

References