Using Eye-tracking to Assess the Relative Importance of Visual and Vestibular Input to Subcortical Motion Processing in the Roll Plane

We study how eye movements reflect brain function. The aim is to develop better diagnostics and treatments for neurological conditions by understanding how vision, balance, and motion perception interact. We have established a method quantifying visual and vestibular contributions to motion perception at the subcortical level, showing how concussion may lead to visual vertigo through increased sensitivity to visual motion.

We will explore how eye movements contract disease progression and treatment response in neurological disorders. We are currently exploring how Galvanic super cortical stimulation might decrease motion sickness and promote gaze stability. To begin, seat the subject securely in the designated chair for all trials.

Adjust the chair position to provide both stability and comfort and to reduce the risk of unwanted head movements or mask slippage. Using hook and loop straps, place and secure the head-mounted eye tracker onto the subject's head to minimize head movement. Confirm that the eye tracker cameras have a clear and unobstructed view of the subject's eyes throughout all movements.

Modify the rotation point height to accommodate individual height differences. Then adjust the rotation point of the mechanical sled so that the axis of whole body rotation is set between the subject's eyes. Firmly secure the head-mounted tracker to the subject's head.

Choose a high contrast visual scene composed of scattered lines or dots centered around a fixation point. Position the fixation point so it aligns directly in front of the subject's eyes, both vertically and horizontally. Eliminate all distracting light sources in the room so that the visual scene is the only source of illumination.

Use a display screen that is large enough to fill the subject's entire visual field. Instruct the subject to keep their gaze fixed on the central point throughout the trial. Then start the eye and head tracking software and present the static visual scene for 10 seconds before beginning any motion.

Between one and two seconds before initiating motion, instruct the subject to keep their eyes wide open. Begin the visual motion by rotating the scene to a fixed amplitude at a predetermined acceleration. Ensure the room is completely dark to remove any visual directional cues.

Secure the subject in the mechanized sled to minimize unintended head or body movements. Inform the subject that the trial is about to start. Then start the eye and head tracking software, allowing a ten-second interval to pass before initiating the movement.

Between one and two seconds before movement onset, instruct the subject to keep their eyes wide open. Activate the mechanical sled to perform a head rotation with the same amplitude and acceleration used for the visual motion stimulation. Use an eye tracking software to analyze the recorded eye tracking videos and extract torsional, vertical, and horizontal eye movements.

Configure and calibrate the pupil tracking function according to the system's guidelines. For torsional response analysis, select two reference points with distinct topographical features on either side of the pupil for each eye, enabling accurate template matching. Run the analysis program to generate positional data over time and export all eye movement data into a separate file.

Now digitize the input data from the eye tracking system, including eye movements, head position, and chair motion. Visually inspect the imported data for each trial to review torsional, vertical, and horizontal eye positions over time, along with head position in the roll plane. Then confirm a stable baseline and expected movement responses across all data streams.

To analyze slow phases, manually trace each torsional slow phase to exclude any remaining confounding data. Evaluate the timing of nystagmus beats by marking the onset of each quick phase and counting the total number of quick phases per trial and subject. Include all slow phase traces from every trial and subject to ensure representative high-powered statistical data.

To evaluate sensory contributions, divide each subject's mean slow phase velocity from visual-only and vestibular-only trials by that from visuo-vestibular trials averaged per acceleration condition. Across all stimulation conditions, visuo-vestibular trials produced the highest torsional slow phase velocities, while visual-only trials resulted in the lowest, validating additive multisensory integration. Torsional velocity increased systematically with higher stimulus acceleration across all modalities, demonstrating acceleration-dependent sensitivity.

Patients showed significantly higher torsional slow phase velocities than controls during both visual and visuo-vestibular stimulation, but not during vestibular-only trials. Eye stimulus gain was highest during visuo-vestibular trials, moderate during vestibular-only, and lowest in visual-only stimulation, confirming modality-specific tracking sensitivity. Relative contribution analysis showed that vestibular input consistently outweighed visual input at all acceleration levels, with the disparity increasing at higher accelerations.

Compared to controls, patients had reduced vestibular contribution and increased visual contribution, indicating altered sensory weighting post-concussion. Nystagmus beat frequency did not differ across groups or modalities, but patients exhibited earlier onset of beats in visual trials, especially at higher accelerations.