Assessing the Reliability of Smooth Pursuit Across Various Neck Postures Using a Custom Ocular Motor Detection System

Overview

Background and Purpose: The assessment of ocular motor control through the manipulation of diverse sensory stimuli can facilitate the diagnosis of dizziness. However, the testing process typically encompasses a single rotational axis. This study aims to investigate the performance of ocular motor control under multi-axis neck rotation postures. To verify the data quality performance and reliability of the custom-made ocular motor features detection system, validate the consistency of system eye movement parameters, and investigate the effect of different neck plane positions on tracking performance. Methods: A total of seven healthy volunteers participated in this study. The study involved ocular motor tracking tasks in different neck positions with all seven subjects, A total of seven subjects were examined, comprising neutral neck posture, left lateral flexion posture, right lateral flexion posture, neck extension posture, neck flexion posture, left neck torsion posture, and right neck torsion posture. The study evaluated eye movement parameters in different planes, including Gain (representing pursuit speed), SPNTD (representing pursuit differences across planes), Accuracy (representing pursuit accuracy), and Latency (representing pursuit latency), as well as the reliability of these parameters. The reliability of the parameters was verified using the intraclass coefficient (ICC). Non-parametric tests (Friedman test) were used to verify the performance of ocular motor tracking in different neck positions, and post-hoc analysis (Wilcoxon sign test) was used to verify statistical differences.

The assessment of ocular motor control through the manipulation of various sensory stimuli helps diagnose dizziness. However, most existing tests focus solely on a single rotational axis. This study aims to evaluate the performance of ocular motor control under multi-axis neck rotation postures. The research objective was integrated the desktop eye tracker with a six-axis Stewart platform to establish an ocular motor features detection system. This system will be used to validate the consistency of the system in ocular motor parameters and explore how different neck plane positions affect tracking performance. Methods: The smooth pursuit task was conducted under seven different neck postures: 1. Neutral neck posture 2. Left lateral flexion posture 3, Right lateral flexion posture 4, Neck extension posture 5\. Neck flexion posture 6. Left neck torsion posture 7. Right neck torsion posture For each posture, participants were seated in a chair with their heads fixed in position to prevent any movement that could stimulate the vestibular system. The Stewart platform was used to control the neck posture of the participant by moving their body, allowing for specific neck rotations or flexions without moving the head. This setup ensured the focus remained solely on the effect of neck posture on smooth pursuit eye movements. Participants were asked to track a light point moving horizontally across the screen in a sinusoidal trajectory. The target's movement was designed based on Simple Harmonic Motion (SHM) to allow for smooth, continuous movement that minimizes abrupt stops, which can trigger saccadic eye movements (quick, jerky movements). SHM was chosen because it better matches the natural movement of smooth pursuit eye tracking. The target speed was set at 30° per second, and the amplitude of the target's movement ranged from 30° to 40°, based on previous studies indicating that these values were optimal for distinguishing between healthy individuals and those with vestibular or cervical dysfunction. The task required participants to follow the moving target as closely and smoothly as possible, while their eye movements were recorded using the Gaze-point 3 (GP3) eye tracker, which sampled at 60Hz. Key Parameters Recorded: 1. Gain: This parameter represents the ratio of eye velocity to target velocity, effectively measuring how well the eyes can match the speed of the moving target. A gain of 1 indicates perfect pursuit, while values less than 1 indicate that the eyes are lagging behind the target. 2. SPNTD: This parameter (Smooth Pursuit Neck Torsion Difference) compares the smooth pursuit performance in neutral neck posture to that in various torsional neck postures. It helps to evaluate the impact of different neck positions on smooth pursuit abilities. 3. Accuracy: This measures how precisely the eyes follow the target. It reflects the angular error between the actual eye position and the position of the moving target at any given time. 4. Latency: This parameter measures the delay in the onset of smooth pursuit after the target begins moving. A lower latency indicates a faster response, which is important for evaluating the ability of participants to start tracking the target quickly and smoothly. Data Analysis: The reliability of the system and the parameters mentioned above were evaluated using the Intraclass Correlation Coefficient (ICC). The ICC was calculated for gain, SPNTD, accuracy, and latency to assess how consistently the participants; smooth pursuit performance could be measured across different neck postures. Additionally, a Friedman test (a non-parametric test) was conducted to determine whether there were statistically significant differences in smooth pursuit performance across the various neck postures. For positions that showed significant differences, Wilcoxon signed-rank tests (a post-hoc analysis) were applied to identify which specific neck postures resulted in significant changes in smooth pursuit performance.