The Bull's Eye Pattern of the Tear Film in Humans during Visual Fixation on En-Face Optical Coherence Tomography

Abstract

The aim of the study was to define and characterize the optical behavior of the tear film during visual fixation in humans on en-face optical coherence tomography (OCT). We included 20 healthy participants, 60% female, aged from 25 to 42 years (33.05 ± 4.97 [mean ± SD]) and ten patients with severe dry eye, 50% female, aged from 26 to 42 years (33.7 ± 5.31). To perform high-resolution tear film imaging, participants were asked to gaze at the internal fixation point in the spectral-domain anterior segment OCT device, and meanwhile scanning session was executed at the following time-points after blinking: at the 2^nd, 3^rd, 4^th, 5^th, and 6^th second. After one hour, OCT imaging was repeated (second session) by a different operator masked to the study to verify the reliability of results. During each measuring session, a pulse oximetry was used for continuously measuring the heart rate and oxygen saturation (SpO₂%). A preliminary experiment was also performed to test the absence of geometric patterns from the anterior surface of a motionless artificial eye. OCT imaging showed a motionless, stable anterior surface of the artificial eye and in dry eye patients. Conversely, in the healthy participants of the study, a bull's eye pattern of the tear film was detected by OCT at the 2^nd, 3^rd, 4^th, 5^th, and 6^th second after blinking, respectively, in 45%, 60%, 45%, 60%, and 40% of OCT scans during the first session, and in 35%, 65%, 65%, 60%, and 35% of cases in the second session. Overall, a total of 200 OCT scans were performed in normal human population. A significant correlation was found between the novel tear film pattern and heart rate during the first and the second session (p < 0.01) in healthy eyes. Conversely, no correlation was revealed with SpO2%. Intraclass correlation (ICC) analysis for OCT imaging of the tear film revealed a statistically significant reproducibility of the results (ICC = 0.838; p < 0.01), indicating the high level of reliability of the method, independently of heart rate and SpO2% variables. There exists a novel, geometric pattern of the tear film during visual fixation detectable by en-face OCT, which is mainly evident as heart rate increases. Its discovery implies in turn the presence of a specific vibration (or imperceptible motion) of the tear film that, at present, is not recognized and corrected by the OCT software (in image postprocessing) unlike other eyeball movements.

Figure 1

A Bull’s Eye Pattern from the tear film surface on en-face OCT imaging. This geometric pattern was characterized by multiple figures, mainly concentric circles, which cyclically reverse their intensity (with different color scale values) from the center towards the periphery. This bull’s eye appearance was detected only in healthy subjects, and never in the context of the cornea (see also Video SI2). (Sectors were as follows: N: nasal; T: temporal; S: superior; I: inferior. Each scan line within the data volume was labeled with a number. Scale bar = 1 mm). Reflectivity of the en-face image [Color bar: intensity = 0 (dark blue) to intensity = 1 (red)].

Figure 2

A regular, tear film pattern (free from geometric figures) by en-face OCT imaging. A reflectivity approximately homogeneous (except for some lines, or corneal reflex artifacts) over the tear film surface was observed in OCT images free from concentric circles. We defined this pattern as regular, stable, or quiescent. This appearance was found both in the artificial eye (top), and in all dry eye patients (bottom). (Sectors were as follows: N: nasal; T: temporal; S: superior; I: inferior. Each scan line within the data volume was labeled with a number. Scale bar = 1 mm). Color bar as in Fig. 1.

Figure 3

En-face OCT imaging of an artificial eye, motionless and dry. En-face OCT scan of an artificial eye. To verify the presence of patterns of geometric figures from the anterior surface of the eye (see text), a “test eye” (normally used for calibration of IOLMaster biometer) was placed on a mechanical support in front of OCT device. A stable anterior surface of the artificial eye was obtained by OCT imaging, confirming the absence of geometric figures or image distortion (please see also Video SI1). (Sectors were as follows: N: nasal; T: temporal; S: superior; I: inferior. Each scan line within the data volume was labeled with a number. Scale bar = 1 mm). Color bar as in Fig. 1.

Figure 4

En-face OCT imaging of an artificial eye with a mucomimetic or lipid layer. A mucomimetic (left) and a lipid-based (middle) artificial tear was separately instilled on two artificial eyes in order to unveil the optical behavior of the mucoaqueous and lipid layer on en-face OCT imaging (e.g. verifying the presence of a hypothetical thin-film interference effect). As a result, a regular pattern with increased reflectivity was detected in both cases, but the bull’s eye appearance has never been detected. The 3-D image of the artificial tear film (right) helps to understand the great ability to reflect the OCT signal from these layers. (Sectors were as follows: N: nasal; T: temporal; S: superior; I: inferior. Each scan line within the data volume was labeled with a number. Scale bar = 1 mm). Color bar as in Fig. 1.

Figure 5

Schematic diagram of coordinate system (x, y) used to describe the interaction between a wavy tear film and light rays. Propagation of light rays (arrows) at the interface between air and tear film depends on whether they travel perpendicular to the boundary between the two media (angle of incidence measured from the normal, θ_i = 0) or with an oblique angle (θ_i ≠ 0). Exact values of refractive indices are assumed to be non influential, for simplicity. Considering the light rays perpendicular to the tear film plane, the movements of the tears, acting as “irregularly-curved surfaces” or irregular lenses (i.e. as an inhomogeneous medium), can lead to changes in direction of propagation of incident light rays depending on whether they strike a point of surface wave “intermediate” (i), or “extreme” (that is the crest (C) or the trough (T), which are respectively the highest and the lowest point of the disturbance in the tear film). In particular, at intermediate points (i) between C and T, there is a change in direction of incident light rays (since the refractive indices of the two media are different) that implies a deviation angle (θ_d) in their propagation (dotted lines): θ_d ≠ 0. Conversely, the light rays that strike the most extreme points of the wave (traveling perpendicular to the boundary between the media) can continue their route beyond the free surface without being deflected (θ_d = 0), with a change only in their speed (dashed lines). Since the waves travel or vibrate through the free surface of the tear film, the propagation angle (θ_d) of incident light rays varies over time and the spatial points placed behind the tear film are alternatively struck by them one after the other, cyclically.