Ultrasound Assessment of Gaze-induced Posterior Eyewall Deformation in Highly Myopic Eyes

Abstract

Purpose: To establish a quantitative metric of posterior eyewall deformability in different directions of gaze in highly myopic eyes with and without posterior staphyloma.

Methods: A prospective study was performed on 53 highly myopic patients (106 eyes). Ultrasound scans were acquired in primary, up, downward, nasal, and temporal gazes. A validated intensity-based segmentation algorithm was used to quantify the posterior eyewall geometry on digitalized B-scan images. Posterior eyewall local curvature (K) and distance (L) to the transducer were calculated. The associations between directions of gaze, axial length (AL), and presence of staphyloma with the K and L parameters were assessed.

Results: A total of 53 participants (106 eyes) were studied. Multivariate regression analysis demonstrated that, after accounting for longer AL, and presence of staphyloma, eccentric gaze was often independently associated with various K and L parameters. Specifically, downward gaze was associated with increased posterior eyewall concavity as reflected in the maximum of K (K_Max) (β = 0.050, P < 0.001) and absolute value of K_Max (β = 0.041, P = 0.011). Both downward gaze and upgaze were independently associated with increase in the derivative of absolute K_Max (which is consistent with more apparent, steeper staphyloma ridges), local K_Max (which detects K_Max at smaller intervals), and K_std (which represents likelihood of staphyloma presence) and decrease in maximum of L (which represents movement of the staphyloma apex) with all P < 0.05. The β coefficients for downward gaze were consistently greater in magnitude compared with those in upgaze. After accounting for AL and presence of staphyloma, horizontal gazes were independently associated only with decrease in the standard deviation of L (which also represents likelihood of staphyloma presence) and maximum of L.

Conclusions: Downward gaze results in a significant increase in posterior eyewall concavity in highly myopic eyes after accounting for AL and staphyloma presence. In comparison with downward gaze, upgaze resulted in a lower magnitude, but significant changes in staphyloma ridge steepness and the likelihood of staphyloma presence. Horizontal gazes seemed to be associated with less posterior eyewall geometric parameters. Studies are required to further assess the association between downward gaze during near work on posterior eyewall concavity and possible effects on myopia development and progression.

High myopia, which can be defined as a refractive error of at least −6.00 diopters (D), is a significant public health issue, particularly in East Asia.^1–3 It is estimated that, by 2050, approximately one-half of the world's population will have myopia and up to 20% of the population with myopia will be highly myopic.⁴ High myopia is associated with pathologic myopia. Pathologic myopia, which can lead to the loss of best-corrected visual acuity, is characterized by excessive axial elongation, which leads to marked scleral thinning, myopic maculopathy, and/or staphyloma formation.^1,3 

Staphyloma is defined as an outpouching of the wall of the eye that has a radius of curvature that is less than the surrounding radius of curvature of the wall of the eye.⁵ Staphylomatous eyes exhibit global and local variations in shape and a greater nonuniformity of the posterior eyewall compared with normal eyes.^6,7 These structural changes have been characterized using imaging modalities such as magnetic resonance imaging (MRI), optical coherence tomography (OCT), and ultrasound examination.^6,8–12 Using ultrasound in real time, our group had previously developed a simple eyewall curvature- and axial length (AL)-based algorithm to automate detection of staphyloma ridge and apex locations that yielded a diagnostic validation performance of 0.897, which was comparable with the diagnostic performance of junior clinicians.¹² Ophthalmic ultrasound examination affords many advantages over other modalities and can (1) image the entire eye dynamically, (2) provide deeper penetration, (3) function in the presence of media opacities such as cornea opacities, cataracts, and vitreous hemorrhage, and (4) generate quantitative morphological data of the posterior eyewall and staphyloma.¹² 

Among the environmental factors that contribute to axial elongation in myopia, near work has been reported to be associated with myopia development and progression.^1,13–15 Near work often involves downward gaze, such as when one reads from a mobile phone or tablet and when one performs desk-based tasks. It has been hypothesized that repeated episodes of downward gaze result in axial elongation, scleral remodeling, and myopia development.¹⁶ In a study using a noncontact optical biometer, Ghosh et al.¹⁷ reported that downward gaze and accommodation was associated with an increase in AL and that there was a greater extent of axial elongation in eyes with a longer AL. Read et al.¹⁸ also used noncontact optical biometry and similarly reported statistically significant axial elongation of less than 10 µm during 6 D of accommodation. Separately, our group assessed the vitreous chamber axial volume of participants using three-dimensional MRI and reported that, although significant gaze-induced globe deformation was noted in all directions of gaze, the vitreous chamber axial volume increased significantly only in downward gaze.¹⁰ 

It would be advantageous to use parameters that characterize three-dimensional structural changes in the posterior eyewall and staphyloma instead of using linear two-dimensional parameters like AL. However, MRI is slower and prone to artifacts. Building on our previous work,¹² we aimed to quantify and compare real-time changes in the posterior eyewall in highly myopic eyes in different directions of gaze using ultrasound examination. The aim of this study was to establish an ultrasound examination-based quantitative metric of posterior eyewall deformability and to assess the extent of posterior eyewall deformation in different directions of gaze in highly myopic eyes with and without posterior staphyloma. 

Methods

This study adhered to the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act of 1996. Ethics approval was obtained from the Institutional Review Board of the Singapore Eye Research Institute and from the Columbia University Irving Medical Center (New York, NY, USA). Written informed consent was obtained from all participants. All participants have a spherical equivalent of at least −6.00 D and a best-corrected visual acuity of at least 6/12 in at least one eye. Patients were excluded if they had a history of prior intraocular surgeries apart from uncomplicated cataract surgeries. 

Every participant was examined comprehensively using a standardized protocol, including a dilated fundus examination, AL measurements (IOLMaster; Carl Zeiss Meditec, Inc., Dublin, CA, USA), swept-source OCT (Atlantis; Topcon, Oakland, NJ, USA), and B-scan ultrasonography with a 10-MHz probe (Quantel Aviso; Rockwall, TX, USA). B-scan images of all participants were screened for the presence of a staphyloma, which was subsequently verified using MRI. 

Ultrasound Protocol

Two independent sequences of ultrasound B-scan data were acquired by an experienced ophthalmic ultrasound imaging researcher (RHS) and a senior fellowship-trained retinal specialist (QVH). The ultrasound probe was softly positioned on the closed eyelid with a coupling gel. Two scan sequences were acquired for each eye: a vertical scan (primary gaze, upgaze, and downward gaze) followed by a horizontal scan (primary gaze, nasal gaze, and temporal gaze) in the anterior-posterior direction. See Figure 1 for representative images. Gain, dynamic range, time–gain compensation, and other data acquisition settings were optimized for each patient. Acquired images were digitized and used for analysis. 

Figure 1.

 

Representative ultrasound B-scan images of eye in horizontal (top) and vertical scans (bottom). The eye is shown in primary gaze (middle), nasal gaze (top right), temporal gaze (top left), upgaze (bottom left), and downward gaze (bottom right).

Staphyloma Detection

On the ultrasound B-scan, the lateral boundaries of the staphyloma (“ridges”) and the staphyloma apex were manually identified in each ultrasound B-scan. The location of a staphyloma was defined as the region bounded by its visible ridges within the plane of the image. 

Image Processing

All processing was performed in MATLAB and Statistics Toolbox Release 2018b (The MathWorks, Inc., Natick, MA, USA). 

The vitreoretinal boundary obtained from the ultrasound B-scan image was used to represent the curvature of the posterior eyewall. The posterior eyewall was automatically detected by an automatic intensity-based segmentation method. Figure 2 illustrates the segmentation of the posterior eyewall and apex in primary and downward gaze. The boundary of the posterior eyewall was defined as the boundary between the black (vitreous) and white (posterior eyewall) pixels, and thus the boundary represents the vitreoretinal interface. After applying a two-dimensional anisotropic diffusion filter to decrease noise while preserving the edges, the image was binarized by an optimally determined threshold (i.e., average image intensity ×4). The noise that remained after thresholding was removed by opening and closing morphological operations. To remove other unwanted segments such as anterior sections, only the largest connected component in the image after thresholding was retained as the posterior eyewall. 

Figure 2.

 

Illustration of a three-dimensional rendering of a pathologic myopia eye with the anterior cornea to the left and posterior staphyloma to the right. The first column from the left depicts the two-dimensional planes in which the ultrasound B-scan images used in this study were obtained, namely, a vertical plane and a horizontal plane that both bisected the eye through the visual axis (red). AL was measured along the visual axis with the IOLMaster (Carl Zeiss Meditec, Inc., Dublin, CA). The second column from the left shows MRI scans done in the vertical plane and horizontal plane (first column). A three-dimensional MRI of a representative pathologic myopia eye with posterior staphyloma is shown in downward gaze in the vertical plane (viewed from below the eye) and primary gaze in the horizontal plane (viewed from temporal to the eye). Their corresponding ultrasound images bisecting the eye through the visual axis are shown in the respective midline vertical plane and midline horizontal plane (third column from left). For these ultrasound images, the autosegmentation lines of the vitreoretinal boundary in the posterior eyewall are tracked in pink and the estimated apex is shown in blue (fourth column).

The posterior eyewall boundary was extracted from the binarized image, and the identified boundary was smoothed by a one-dimensional median filter. A cubic smoothing spline was then fitted to the detected boundary to further smooth the detected surface. Parts of the detected boundary that extended past the B-scan image and into the background were truncated, where the B-scan region had been automatically detected via thresholding. Subsequently, the curvature K (in mm^–1) was calculated from the segmented boundary at each specific point (x_n, z_n) on the detected boundary. Twenty percent of the data length from each side from all images were uniformly cropped out to remove artifacts owing to decreased signal at the edges of the image (i.e., 40% of the data were cropped out). The remaining central 60% of the segmentation was used for further analysis. 

Assessment of Posterior Eyewall Geometry

K (posterior eyewall curvature) and L (distance from the transducer) parameters were calculated for each B-scan. These include maximum of K (K_Max), maximum of the absolute value of K (|K|_Max), maximum of the absolute value of derivative in K (denoted as |K_x|_Max), maximum local standard deviation of K (local K_Max), standard deviation of K (K_std), standard deviation of L (L_std), and maximum of L (L_Max). These K and L parameters were previously validated as part of a fully automated classification algorithm that detects the posterior eyewall and quantifies its shape.¹² Conventionally, R is used to denote curvature and r is used to denote the radius of curvature in mm¹⁹; however, we have used K to indicate posterior eyewall curvature to maintain a consistent nomenclature with our previous work.¹² K_Max is particularly relevant to our study of myopic eyes, because it directly quantifies the maximum concavity of the posterior eye wall, and in pathologic myopia eyes, specifically the concavity of the most prominent staphyloma. See Table 1 for the mathematical definitions and clinical interpretations of these parameters. 

Table 1.

 

Definitions of Parameters

 

K can be obtained by calculating the following at each point (x_n, z_n) on the posterior eyewall:  

\begin{eqnarray} K = {\rm{\ }}\frac{{\frac{{{d^2}z}}{{d{x^2}}}}}{{{{\left( {1 + {{\left( {\frac{{dz}}{{dx}}} \right)}^2}} \right)}^{\frac{3}{2}}}}},\quad \end{eqnarray}

(1)

where x and z are the lateral and axial dimensions, respectively, and \(\frac{{dz}}{{dx}}\) and \(\frac{{{d^2}z}}{{d{x^2}}}\) are the first and second derivatives. 

In the discrete domain, K at point (x_n, z_n) can be written as the following:  

\begin{eqnarray} {K_n}{\rm{\ }} = {\rm{\ }}\frac{{z_n^{{\rm{^{\prime\prime}}}}}}{{{{\left( {1 + {{\left( {z_n^{\rm{^{\prime}}}} \right)}^2}} \right)}^{\frac{3}{2}}}}},\quad \end{eqnarray}

(2)

where  

\begin{eqnarray} z_n^{\rm{^{\prime}}}{\rm{\ }} = {\rm{\ }}\frac{{{z_{n + 1}} - {z_n}}}{{{x_{n + 1}} - {x_n}}},\quad \end{eqnarray}

(3)

 

\begin{eqnarray} z_n^{{\rm{^{\prime\prime}}}}{\rm{\ }} = {\rm{\ }}\frac{{z{{\rm{^{\prime}}}_{n + 1}} - z{{\rm{^{\prime}}}_n}}}{{{x_{n + 1}} - {x_n}}}.\quad \end{eqnarray}

(4)

 

The magnitude of K indicates the curvature within the analysis range, and the sign of K indicates the direction of the curve (positive if convex toward the lens, negative if convex toward the optic nerve). 

Assessment of Staphyloma Apex Location

The location of the staphyloma apex was also assessed using the detected posterior eyewall for the eyes with a staphyloma. It was empirically observed that the apex of the staphyloma tends to be located the furthest from the transducer. Thus, the apex was defined to be located at the position of the maximum distance from the center point of the transducer (x_T, z_T). Note that the center point of the transducer was defined as the mechanical pivot point of the sector (i.e., x_T was defined as the midpoint in the x axis, z_T was defined as 0 along the z axis). The L at each point on the segmented posterior eyewall was calculated as:  

\begin{eqnarray} {L_n}{\rm{\ }} = {\rm{\ }}\sqrt {{{\left( {{x_T} - {x_n}} \right)}^2} + {{\left( {{z_T} - {z_n}} \right)}^2}}.\quad \end{eqnarray}

(5)

 

Thereafter, the apex (x_apex, z_apex) was defined as the index that maximizes L. The error of the apex localization was calculated as the lateral distance between the automatically determined apex location and the clinically defined apex location. 

L _std correlates with posterior eyewall deformation because it measures the standard deviation of L within the same ultrasound image. A smaller L_std is expected if the posterior eyewall is a uniform arc, whereas a larger L_std is expected if the posterior eyewall has a prolate shape or a nonuniform, variable curve and aberrant shape (e.g., in the presence of staphyloma[s]). L_Max, which measures the longest distance from the posterior eyewall to the transducer, is similarly expected to be greater in prolate eyes and more so in prolate eyes with staphyloma. Although the value of L_Max may not directly represent deformation, L_Max may readily change in value in different directions of gaze if the posterior eye curvature changes with deformation. 

Statistical Analysis

Statistical analysis was performed using STATA (Version 16.1, StataCorp LLC, College Station, TX, USA). Generalized estimating equation (GEE) linear regression was first used in the univariable analysis to examine crude associations between direction of gaze, the presence of staphyloma, and the AL with the posterior eyewall geometry parameters as defined in Table 1. A GEE was used to consider the possible correlation between the two eyes of an individual in the regression analyses. This process was followed by manual stepwise multivariable GEE linear regression analyses, with adjustment for confounders. 

The normality of the distributions of parameters was assessed using graphical (histogram) and statistical (Shapiro–Wilk) means. All continuous variables were assessed to be normally distributed. Categorical variables were represented as counts and percentages. All P values are two-tailed. A P value of <0.05 was defined as significant. For all models, the β coefficients and 95% confidence intervals (CIs) were reported. 

Results

A total of 53 participants (106 eyes) were studied, including 19 East Asians, 29 Caucasians, and 5 African Americans. The mean age was 59.4 ± 12.7 years (range, 30–83 years). The mean AL was 30.7 ± 3.0 mm (range, 27.0–39.3 mm). 

Table 2 illustrates the K and L parameters in different directions of gaze (reported as mean ± SD for 106 eyes). Specifically, the significance of each gaze direction for a parameter was evaluated based on a univariate GEE model that takes into account correlation between each individual's left and right eyes. The changes in K_Max (degree of posterior eyewall concavity) and local K_Max (K_Max as measured at specified intervals) were observed only in downward gaze (0.085 ± 0.115 and 0.115 ± 0.052, respectively, which were significantly greater than that in primary gaze: 0.042 ± 0.069 [P = 0.001] and 0.096 ± 0.046 [P = 0.004], respectively). |K|_Max (overall degree of posterior eyewall curvature change combined, regardless of direction, including both concavity and convexity) was not associated with either vertical or horizontal eye movements. 

Table 2.

 

Posterior Eyewall Curvature in Different Directions of Gaze

 

The |K_x|_Max (derivative of absolute value of the rate of spatial change in curvature, indicative of steepness of staphyloma ridges), showed changes in both downward gaze and upgaze (downward, 0.426 ± 0.235; up, 0.395 ± 0.191 vs. primary gaze, 0.345 ± 0.194 [P = 0.007 and 0.030, respectively]). K_std (nonuniformity along the curve over the entire posterior eyewall boundary that represents likelihood of staphyloma presence along the posterior eyewall) also showed changes in both downward gaze and upgaze (downward, 0.099 ± 0.042; up, 0.091 ± 0.038 vs. primary gaze, 0.083 ± 0.037 [P = 0.001 and 0.040, respectively]). 

The L_std (variation of the distance from the posterior eyewall to transducer, which similarly represents likelihood of staphyloma presence along the posterior eyewall) was associated with significant decrease in only temporal gaze (0.576 ± 0.350 vs. primary gaze 0.689 ± 0.504 [P = 0.028]). Not surprisingly, the L_Max (mathematically the longest distance from the posterior eyewall to the transducer, which represents the location of the staphyloma apex) was associated with both horizontal and vertical eye movements, with P values all <0.01. This finding is consistent with our prior MRI study that reported that three-dimensional gaze-induced eye deformations occurred in all four eccentric gazes.¹⁰ 

Subsequently, through multivariate regression analyses, we assessed associations between posterior eyewall curvature and directions of gaze, adjusting for AL and the presence of staphyloma (Table 3). Overall, we found eccentric gaze was often independently associated with various K and L parameters, with the greatest magnitude of change found with downward gaze, compared with upgaze and horizontal eye movements. 

Table 3.

 

Associations With Posterior Eyewall Curvature

 

Compared with primary gaze, downward gaze was associated with K_Max (β = 0.050; 95% CI, 0.023–0.076), |K|_Max (β = 0.041; 95% CI, 0.009–0.074), |K_x|_Max (β = 0.093; 95% CI, 0.034–0.152), local K_Max (β = 0.023; 95% CI, 0.011–0.035), K_std (β = 0.020; 95% CI, 0.010–0.029), and L_Max (β = −3.358; 95% CI, −5.386 to −1.330). Although downward was the only gaze direction associated with increased posterior eyewall concavity (as reflected in K_Max and |K|_Max), both downward gaze and upgaze were independently associated with increase in the derivative of |K|_Max (consistent with steeper staphyloma ridges), local K_Max, and K_std (which represents the likelihood of staphyloma presence) and decrease in L_Max (which represents movement of the staphyloma apex) with all P < 0.05. The β coefficients for downward gaze appeared consistently greater in magnitude compared with those in upgaze. Specifically, a separate analysis (Table 4) shows that these changes in downward gaze were significantly greater compared with that in upgaze for K_Max, |K|_Max, local K_Max, K_std, and L_Max. There were no significant differences in the changes between downward gaze and upgaze for |K_x|_Max and L_std. 

Table 4.

 

Associations With Posterior Eyewall Curvature for Vertical Gaze (With Upgaze as the Reference Group for Direction of Gaze)

 

After accounting for AL and the presence of staphyloma, horizontal gazes were independently associated only with decrease in L_std (which also represents the likelihood of staphyloma presence) and L_Max. Specifically, temporal gaze (β = −0.115; 95% CI, −0.207 to −0.023) was associated with a smaller L_std, and both nasal gaze (β = −3.235; 95% CI, −4.970 to −1.499) and temporal gaze (β = −2.747; 95% CI, −4.326 to −1.167) were associated with a smaller L_Max. Horizontal gaze was not associated with the K parameters. 

After adjustment for direction of gaze and the presence of a staphyloma, AL still exhibited a positive relationship with all K and L parameters (P < 0.05), except for K_Max (degree of concavity) and |K_x|_Max (steepness of staphyloma ridge). This finding indicated a greater tendency for a more altered posterior eyewall geometry in participants with a longer eye. 

After adjustment for direction of gaze and AL, the presence of a staphyloma still exhibited a positive relationship with all K and L parameters (P < 0.05), except for L_Max and horizontal image |K_x|_Max. Not surprisingly, this result indicated a greater tendency for further altered posterior eyewall deformation in participants with a staphyloma. 

Discussion

This study has established a quantitative metric of posterior eyewall deformability using ultrasound examination and demonstrated that downward gaze resulted in significant deformation of the posterior eyewall in a cohort of highly myopic patients. A longer AL and the presence of a staphyloma were also independently associated with greater deformation of the posterior eyewall. 

Previous studies have reported changes in AL with downward gaze, accommodation, and/or convergence using a variety of modalities.^6,8–12 We sought to assess and compare three-dimensional and complex posterior eyewall changes in different directions of gaze with ultrasound examination, which is readily available in the clinic. To this end, we used ultrasound-derived K and L parameters to quantify the posterior eyewall curvature changes and the distance from the posterior eyewall to transducer in different directions of gaze. 

We report in multivariate regression analysis that controls for AL and staphyloma presence, that downward gaze was significantly associated particularly with the K parameters (K_Max, |K|_Max, |K_x|_Max, local K_Max, K_std, and L_Max), which indicated a greater posterior eyewall concavity, staphyloma ridge steepness, and likelihood of staphyloma and change in the staphyloma apex location. These changes in downward gaze were significantly greater compared with that in upgaze for K_Max, |K|_Max, local K_Max, K_std, and L_Max. Among the K parameters, K_Max is particularly important as it directly quantifies the degree of concavity of a staphyloma in pathologic myopia eyes. In contrast, although upgaze was significant for several K and L parameters, the β coefficients were generally smaller than that in downward gaze except for |K_x|_Max and L_std. Horizontal gaze generally resulted in nonsignificant posterior eyewall deformation and was associated only with L_std in temporal gaze and L_Max in nasal and temporal gaze, and with none of the K parameters. Therefore, the extraocular muscles involved in downward gaze may exert a disproportionate and asymmetric stress pattern that deforms the posterior eyewall to a greater extent compared with those that are involved in other directions of gaze. This finding may be attributed to asymmetric insertions and attachments of extraocular muscles and the different positions of the muscle pulleys, which result in differing vector forces and varying extents of posterior eyewall deformation in different directions of gaze. Although the superior oblique and the inferior rectus are used on downward gaze, the inferior oblique and superior rectus are used in upgaze. 

This study has provided evidence that the extraocular muscles can transmit forces to the globe of highly myopic eyes, which results in posterior eyewall deformation, particularly in downward gaze. A relevant question is whether downward gaze in near work results in myopia development and progression. Regarding this, although near work has been reported as a risk factor for myopia development and progression, results of cross-sectional and longitudinal studies have also been inconsistent.^20,21 However, assuming near work does play a role and if we were to hypothesize, given the greater extent of posterior eyewall deformation in downward gaze, a plausible mechanism would involve repeated and prolonged stress that is exerted by the extraocular muscles during near work and downward gaze that leads to myopia development and progression. Grytz et al.¹⁶ had previously postulated that the adaptation of scleral tissue in response to mechanical stress could result in the scleral remodeling mechanism underlying myopia development. Highly myopic eyes typically exhibit overall thinning, localized ectasia, and altered biomechanical properties of the sclera.^17,22 Furthermore, studies that have assessed the biomechanical properties of the sclera in animal models have described that the magnitude of scleral deformation increases with time if a constant pressure is applied to the globe.^23,24 Additionally, small temporary IOP elevations in different directions of gaze have been associated with an increase in AL in young adult participants.¹⁸ However, given the cross-sectional nature of our study, which precludes the inference of a temporal or sequential relationship, and that our cohort were all highly myopic eyes, our study is unable to conclusively provide evidence for the association between near work and myopia development and progression. Studies need to be performed to further assess this association. 

In addition, a longer AL and staphyloma were also independently associated with greater deformation of the posterior eyewall. Presumably, both result in stretching and weakening of the posterior eyewall, increasing its propensity to be deformed, especially during downward gaze. Supporting evidence from animal studies have also indicated that, when compared with emmetropic eyes, myopic eyes demonstrate greater scleral creep extensions in response to sustained strain and are more vulnerable to the stretching influence of mechanical stress.^25,26 These changes may be present even before the onset of scleral thinning and staphyloma formation in pathologic myopia.^27,28 

Furthermore, the posterior pole is particularly at risk of pathology arising from the mechanical stress of extraocular muscles. The scleral stiffness of the posterior pole is approximately 60% of that of the anterior pole.²⁹ Owing to their attachment locations at the posterior globe, oblique muscles may produce significant localized stress and contribute to axial elongation. Moreover, structural pathological changes in myopia like staphyloma formation have a propensity to occur in the posterior pole.³⁰ 

The strength of this study is the use of ultrasound-derived K and L parameters to quantify the shape and extent of the posterior eyewall deformation in different directions of gaze. We had previously validated the use of these K and L parameters as part of a fully automated classification that detects the posterior eyewall and quantifies its shape. Ultrasound examination also has many advantages over OCT or MRI as well, including ease of access, greater penetration depth, wider imaging area, full automation, and accurate localization of staphyloma apex location. 

The weakness of this study includes its relatively small sample size. The K and L parameters were derived from performing and grading the ultrasound examination, which were operator dependent. A potential issue is whether the ultrasound B-scans were performed reproducibly on the same section in each direction of gaze. Although ultrasound B-scan image registration is not possible, as with OCT, we had minimized imperfect registration by tracking the optic nerve as a fiducial point. Moreover, ultrasound scans were all performed by a single, experienced operator (RHS) who had included the optic nerve in all scans as a landmark to gauge the position. Furthermore, three-dimensional MRI scans were also performed separately, which helped to ensure that the ultrasound B-scan was performed on the same section in each direction of gaze. The study was performed on highly myopic eyes, and the relationships between direction of gaze and posterior eyewall curvature may not necessarily apply to eyes of lower myopia levels. In addition, the changes in the K and L parameters were small and reflect the short-term and small influence of shifts in gaze direction on the posterior eyewall curvature. Further research is required to understand the implications of these short-term findings for the long-term myopia development and progression. 

In summary, downward gaze, a longer AL, and the presence of a staphyloma were independently associated with significant deformation of the posterior eyewall in highly myopic eyes. Studies need to be performed to further assess the association between near work and myopia development and progression. 

Acknowledgments

Funded by grants from the National Medical Research Council (MOH-000531-00; MOH-001103-00, to QVH), the SERI-Lee Foundation (LF0621-1, to QVH) the Lee Foundation (TLF 1021-3; TLF 0322-8, to QVH), the SingHealth Foundation-SNEC (R1499/82/2017 to QVH), the Macula Foundation Inc., New York, New York (to LAY), NIH grant P30 EY019007, and an unrestricted grant to the Columbia Department of Ophthalmology from Research to Prevent Blindness. The sponsors or funding organizations had no role in the design or conduct of this research. 

Disclosure: K.X. Cheong, None; S.Y. Lim, None; Y.S. Dan, None; R.H. Silverman, None; S. Chang, None; L.A. Yannuzzi, None; K.B. Freund, None; K. Ito, None; Q.V. Hoang, None