Comparison of Three Gaze-position Calibration Techniques in First Purkinje Image–based Eye Trackers

Abstract

This study highlights potential differences that can arise in gaze position estimates from 1st Purkinje image-based eye trackers based on how individual Hirschberg ratios are calculated. Purpose: To evaluate the accuracy and repeatability of eccentric-viewing, prism-based and theoretical techniques that are routinely used to calibrate Hirschberg Ratio (HR) in 1 st Purkinje image-based eye trackers. Methods: Hirschberg ratios of 28 UK participants (18-40 years) were obtained using the PlusOptix PowerRef3 photorefractor and eye tracker. In the gold-standard eccentric viewing technique, participants viewed eccentric targets (±12°, 4° steps) at 2m. In the prism-based technique, 4-16∆D base-out and base-in prisms were placed in 4∆D steps before an eye occluded with an infrared filter; the fellow eye fixated a target at 1m. Each participant's HR was calculated as the slope of the linear regression of the shift in Purkinje image, relative to the pupil centre, for each target eccentricity or induced prism power. Theoretical HR was calculated from the participant's corneal curvature and anterior chamber depth measures. Data collection was repeated on another visit using all three techniques to assess repeatability. Data were also obtained from an Indian cohort (n=30,18-40 years) using similar protocols. Results: HR ranged from 10.61-14.63°/mm (median: 11.90°/mm) in the eccentric viewing technique. The prism-based and theoretical techniques demonstrated inaccuracies of 12% and 4% relative to the eccentric viewing technique. The 95% limits of agreement of intra-subject variability ranged from ±2.00°/mm, ±0.40°/mm, and ±0.30°/mm for the prism-based, eccentric viewing and theoretical techniques (P>.05). Intraclass correlation coefficients (95%CI) were 0.99 (0.98-1.00) for eccentric, 0.99 (0.99-1.00) for theoretical and 0.88 (0.74-0.94) for prism-based techniques. Similar results were found for the Indian cohort. Conclusion: The prism-based and theoretical techniques both demonstrated relative inaccuracies in measures of Hirschberg ratio, compared to the eccentric viewing technique. The prism-based technique exhibited the poorest repeatability.

Spasm of Near Reflex: Objective Assessment of the Near-Triad

“…The inbuilt value of Hirschberg ratio used by the photorefractor does not show any significant biases and therefore no further scaling of the raw gaze position or convergence data were performed. 19 Pupil diameter is calculated by the photorefractor for each eye, again in-sync with accommodation and gaze position, by detecting the pupil edges using inbuilt image processing algorithms.…”

“…18 Because individual calibration was not possible here because of the subject's unsteady manifest refraction, the raw data were scaled using the average calibration factor for subjects of Indian ethnicity from an earlier study by Sravani et al 18 The individual eye's gaze position is recorded by the photorefractor by tracking the relative separation between the first Purkinje image and the entrance pupil center and applying an in-built Hirschberg ratio (11.8°/mm) to convert this separation into angular units of degrees. 19 Binocular convergence is then derived by taking the difference between the two eyes' gaze positions at each frame. The inbuilt value of Hirschberg ratio used by the photorefractor does not show any significant biases and therefore no further scaling of the raw gaze position or convergence data were performed.…”

Application and repeatability of ocular biometric measures for gaze position calibration in children

“…Some eye trackers use a population average HR (e.g., 11.8 °/mm in the case of the PowerRef 3) to obtain estimates of gaze position. However, significant inter‐subject variability has been reported in the literature, with HR values ranging from 7 to 16 °/mm or (12 to 28 ∆/mm), 7,9,10 which could limit the accuracy of gaze position estimates when such a population‐average is applied. Consequently, individual calibration prior to measurement of gaze position has been recommended 10 .…”

“…Consequently, individual calibration prior to measurement of gaze position has been recommended 10 . Eccentric‐viewing, 9,10 prism‐based 4,7 and theoretical 7,10,11 techniques are the main procedures described and employed for gaze position calibration. With the eccentric‐viewing and prism‐based techniques, PI displacement is induced either by asking subjects to fixate on a series of targets placed at known angular eccentricities or by using prisms.…”

“…With the eccentric‐viewing and prism‐based techniques, PI displacement is induced either by asking subjects to fixate on a series of targets placed at known angular eccentricities or by using prisms. The reciprocal of the linear regression slope for PI displacement versus target eccentricities or prism power provides the HR in degrees or prism dioptres per millimetre 7,10 . However, of particular interest here is the so‐called theoretical technique, which involves an individual's ocular biometric measurements, specifically, the anterior corneal curvature (CC) and the anterior chamber depth (ACD), to estimate the Hirschberg ratio based on a geometric model described by Brodie 8 .…”