Video Head Impulse Test (vHIT)

Introduction to the Video Head Impulse Test (vHIT)

The Video Head Impulse Test, widely known as vHIT, was first described by Halmagyi and Curthoys in 1988 as a bedside clinical test for evaluating the vestibulo-ocular reflex (VOR). The procedure involves monitoring a patient's eye movements while they maintain fixation on a stationary target. During this fixation, the examiner moves the patient's head briskly in the plane of the lateral semicircular canals. While the test is traditionally performed in the lateral plane, it can also be conducted in the planes of the vertical semicircular canals, specifically the right anterior-left posterior (RALP) or the left anterior-right posterior (LARP) canal pairs, although these vertical orientations are considered technically more challenging.

The Bedside Head Impulse Test (HIT) Procedure

In the standard bedside procedure, the examiner stands directly in front of the patient and holds the patient's head firmly. The patient is instructed to maintain a steady gaze at a target located straight ahead, which is typically the examiner’s nose. The examiner then rotates the patient’s head to the right or left unexpectedly using small-amplitude, high-velocity, and high-acceleration movements known as head impulses. In a healthy individual with a normal VOR, the eyes can maintain a steady gaze on the target despite the quick head movement. However, patients with a deficient VOR are unable to keep their eyes on the target during these high-velocity turns. Consequently, their eyes move with the head and away from the target, requiring the generation of compensatory "catch-up" saccades to return the gaze to the target after the impulse is delivered toward the damaged side.

Limitations of the Bedside Procedure

Early clinical experience with the bedside head impulse test yielded mixed results, and the findings did not always correlate with caloric test results. One major limitation is that the bedside test is inherently subjective; it relies entirely on the examiner's visual skill and experience to detect catch-up saccades. Furthermore, the examiner has no immediate feedback regarding the accuracy or the velocity of the head impulses delivered. To overcome these limitations and improve clinical utility, objective methods were developed. The scleral search coil method became the gold standard for accurate measurement of high-velocity eye movements, but its routine clinical use was found to be cumbersome. This led to the development of the video head impulse test, which utilizes high-speed cameras to analyze video images of eye movements, providing a more practical and objective alternative for clinical settings.

Physiological Mechanisms of the Normal Response

The physiological basis for the HIT lies in the asymmetry between the excitatory and inhibitory neural responses of the semicircular canals. Each canal has a tonic firing level of approximately $100\,\text{spikes/sec}$. During excitation, this rate can increase to a maximum of approximately $400\,\text{spikes/sec}$, whereas during inhibition, the firing rate can drop to a minimum of $0\,\text{spikes/sec}$. For natural, slow head movements, neural firing rate changes are proportional to head velocity, and the responses from the right and left sides remain symmetrical. The oculomotor system receives a signal based on the difference between these right and left neural responses. In this state, the VOR Gain, calculated as $\text{Eye movement} / \text{Head movement}$, is equal to $1$.

During more intense head impulses, the inhibitory neural responses saturate quickly at zero, while the excitatory responses remain proportional to the head velocity. Despite this asymmetry, the overall input to the oculomotor system and the resulting eye movements remain symmetrical for right-left head impulses. During these impulses, the responses are mediated primarily by one labyrinth. In these normal conditions, the VOR Gain remains approximately $1$, though it may decrease slightly as head velocity increases.

Pathophysiology of Vestibular Lesions

In the case of a right vestibular lesion, head impulses directed toward the side of the lesion result in neural input to the oculomotor system that is no longer proportional to head velocity. The resulting eye velocity fails to match the head velocity, causing the eyes to fall short of the target. Under these conditions, the VOR Gain is significantly less than $1$ (<<1) and decreases rapidly as head velocity increases. When head impulses are delivered away from the side of the lesion, neural input is still reduced but to a lesser extent. The eye velocity is closer to the target velocity but still falls somewhat short. In this scenario, the VOR Gain is less than $1$ (<1) and decreases with increasing head velocity, though not as rapidly as seen with impulses toward the lesioned side.

Mechanics and Classification of Catch-Up Saccades

When VOR eye movements are insufficient to keep the eyes on the target, the disparity between head and eye positions triggers catch-up saccades. These saccades involve cortical processing, meaning their latency (approximately $100\,\text{msec}$) is significantly longer than that of VOR eye movements (less than $10\,\text{msec}$). Catch-up saccades are classified into two types: overt and covert. Overt saccades occur after the head impulse has finished; they are visible to the naked eye and can be detected by an experienced examiner during a bedside test. Covert saccades occur during the head impulse itself and are practically impossible to detect without specialized equipment like vHIT.

The emergence of covert saccades depends on the time required to predict the head-eye position difference and the time to initiate the saccade. If both these times are decreased, the patient exhibits covert saccades. Conversely, an increase in either the prediction time or the initiation time leads to overt saccades. Because head impulses are unpredictable, covert saccades may not reposition the eyes perfectly on the target, often necessitating a smaller secondary catch-up saccade—typically an overt one—to reach the target exactly. These secondary saccades usually share the same saccadic latency as the primary covert saccade.

Technical Advantages and Equipment Requirements of vHIT

The vHIT offers several advantages, including the ability to identify covert saccades, validation that impulses are performed correctly, and high sensitivity and specificity (95%). It reduces false negatives by identifying truly abnormal patients as such and provides objective analysis based on normative data. Patients often find vHIT more comfortable as traditional visual observation often requires larger head thrusts.

The system requires high-speed ($\approx 250\,\text{frames/sec}$) high-resolution cameras to distinguish between slow and fast movements. The accuracy of head movement sensors, usually embedded in goggles, should ideally be verified against sensors firmly affixed to the head, such as a bite-bar. The fit of the apparatus is critical to accuracy. The device must provide feedback to the examiner to ensure impulses are delivered at appropriate velocities.

Clinical Testing Procedures and Setup

To perform the test, the patient is seated 1 meter from a fixation dot. The goggles must be fitted tightly to prevent slippage. The examiner must click on the pupil to center the green box in the Region of Interest (ROI). Pupil detection is adjusted using auto threshold in grayscale or pupil location mode, ensuring the crosshair locks onto the pupil. The smallest possible pupil size is ideal, so the lights in the room should be kept on. Calibration involves turning on lasers and centering the target between laser points. The patient follows these laser points with their eyes only while the examiner selects "Run" and then "Accept" once calibrated.

Testing begins with the lateral plane. The examiner places hands on top of the patient's head, avoiding the goggles or strap. The patient stares at the fixation dot while the examiner moves the head quickly and unpredictably in small amplitudes of approximately $15^\circ$ to the left or right. Algorithms reject poor impulses, and the examiner monitors the Real Time Window and Training Curve Window. The "sweet spot" for velocity is between $100^\circ/\text{sec}$ and $250^\circ/\text{sec}$.

For vertical plane testing (RALP and LARP), the examiner navigates back to the Test Setup tab. The patient's head is turned $30^\circ$ to $40^\circ$ to the side—counter-clockwise for RALP and clockwise for LARP. The ROI is adjusted, but re-calibration is generally not required before beginning the test.

Criteria for Abnormality and Interpretation

Abnormality is determined by the presence of catch-up saccades (overt and covert), VOR gain values, and gain asymmetry. Abnormal catch-up saccades typically occur for almost all impulses to at least one side and move in the same direction as VOR eye movements. They fall within a specific time window: approximately $100\,\text{msec}$ from the onset of head movement to $100\,\text{msec}$ after the end of the movement. Covert saccades specifically occur from approximately $70\,\text{msec}$ to the end of head motion, while overt saccades occur within approximately $200\,\text{msec}$ after head motion ends. Saccades with velocities significantly smaller than the peak head velocity are not usually considered abnormal; the velocity of an abnormal catch-up saccade is typically comparable to the head velocity.

VOR gain measures the relationship between slow eye movements and head movements, and it can be defined based on position, velocity, or acceleration. VOR gain velocity is often calculated at fixed intervals ($40$, $60$, or $80\,\text{msec}$) after the impulse onset. Position-based VOR gain is determined by the area under the head and eye velocity curves, provided covert saccades are removed first. Normal lateral gain is > 0.8, while normal LARP/RALP gain is > 0.7. Significant gain asymmetry often accompanies abnormal catch-up saccades.

Interaction Between Oculomotor and VOR Mechanisms

The systems contributing to head impulse responses change based on head velocity. For lateral vHIT:

• Responses below $500^\circ/\text{sec}$ (per page 5) or $1000^\circ/\text{sec}$ (per page 7) are mediated by the oculomotor system (smooth pursuit/tracking).

• Both oculomotor and VOR mechanisms contribute between $50\,\text{and}\,1000^\circ/\text{sec}$.

• Responses above $100^\circ/\text{sec}$ (per page 5) or $1000^\circ/\text{sec}$ (per page 7) are mediated entirely by the VOR. Note: The transcript contains conflicting values for these thresholds.

For patient safety, the upper limit for lateral testing is $250\,\text{to}\,3000^\circ/\text{sec}$, with caution advised above $2500^\circ/\text{sec}$. For RALP and LARP, the lower limit for VOR mediation is $500^\circ/\text{sec}$, and the upper safety limit is $2000^\circ/\text{sec}$.

Patterns in Vestibular Pathology

In a normal test, VOR gain is approximately $1$ and may decline slightly as head velocity increases. In a unilateral lesion, VOR gain is reduced for impulses in both directions but is significantly lower for impulses toward the side of the lesion and declines rapidly with velocity. The presence of consistent catch-up saccades on one side confirms a unilateral lesion, even if gain is technically within normal limits. In bilateral lesions, catch-up saccades are present on both sides and gain asymmetry is absent. Total loss results in a gain close to zero. Bilateral unequal loss may resemble a unilateral deficit but can be differentiated by the shape of the VOR eye velocities.

Spontaneous nystagmus appears as spikes in the eye velocity tracing. Unlike catch-up saccades, nystagmus spikes can occur before or after the head impulse, and their velocity is usually much smaller than that of catch-up saccades. In an acute peripheral vestibular lesion, if nystagmus beats away from the lesion, spikes appear in the direction opposite to eye movements when the head is turned toward the intact side.

Clinical Applications and Diagnostic Summary

The vHIT is a versatile tool with numerous clinical applications. It measures absolute responses from each ear, allowing for the detection of partial or complete bilateral loss and hyperactive responses. It is the first test capable of identifying isolated abnormalities in vertical canals and specific branches of the vestibular nerve. vHIT is essential in differentiating cerebellar strokes from peripheral vestibular lesions in acute vertigo cases. It is used for serial monitoring, such as tracking Gentamycin therapy for Meniere’s or vestibulotoxicity. The test is cost-effective, portable, well-tolerated by patients, and can be modified for children. It serves as an objective replacement for rotation chair testing in patients with bilateral caloric weakness.