Biometry-Based Technique for Determining the Anterior Scleral Thickness: Validation Using Optical Coherence Tomography Landmarks

Abstract

Purpose: Considering the potential role of anterior scleral thickness (AST) in myopia and the ubiquitous use of optical biometers, we applied and validated a biometry-based technique for estimating AST using optical coherence tomography (OCT) landmarks.

Methods: The AST was determined across four meridians in 62 participants (aged 20–37 years) with a swept-source OCT and a noncontact optical biometer at a mean ± SD distance of 3.13 ± 0.88 mm from the limbus. The biometer's graticule was focused and aligned with the anterior scleral reflex, which led to the generation of four prominent A-scan peaks: P1 (anterior bulbar conjunctiva), P2 (anterior episclera), P3 (anterior margin of anterior sclera), and P4 (posterior margin of anterior sclera), which were analyzed and compared with the corresponding OCT landmarks to determine tissue thickness.

Results: The AST measurements between biometer and OCT correlated for all meridians (r ≥ 0.70, overall r = 0.82; coefficient of variation [CV], 9%–12%; P < 0.01). The mean difference ± SD between two instruments for overall AST measures was 3 ± 2.8 µm (range, −18 to +16 µm; lower limits of agreement, −89 to +83 µm; P = 0.23) across all meridians. The mean ± SE AST with both instruments was found to be thickest at the inferior (562 ± 7 µm and 578 ± 7 µm) and thinnest at the superior (451 ± 7 µm and 433 ± 6 µm) meridian. The biometer demonstrated good intrasession (CV, 8.4%–9.6%) and intersession (CV, 7.9%–13.3%) repeatability for AST measurements across all meridians.

Conclusions: The noncontact optical biometer, which is typically used to determine axial length, is capable of accurately estimating AST based on OCT landmarks.

Translational Relevance: The high-resolution optical biometers can demonstrate wider application in the field of myopia research and practice to determine AST.

Introduction

Various ocular expansion models such as posterior pole,¹ equatorial,² global,² axial,³ and asymmetrical⁴ models are linked with myopia development and progression.⁵ While there is evidence supporting the hypothesis that ocular expansion in myopic eyes is associated with thinning of various posterior ocular coats (primarily at the level of retina, choroid, and sclera),^6–14 recent studies also indicated thinning of sclera in the anterior regions of the myopic eyes.^4,15 

With the recent advancement in the medical technology, optical coherence tomography (OCT), which is extensively used in imaging posterior ocular structures,^8,16–21 is now also being used for obtaining high-resolution in vivo images of various anterior ocular structures.^4,15,22–31 Buckhurst et al.²² first reported the in vivo measurement of anterior scleral thickness (AST) with anterior segment OCT (AS-OCT) across different ocular meridians in different ethnicities and found that AST was thickest in the inferior (806 ± 60 µm) and thinnest in the superior-nasal (662 ± 57 µm) meridians. A similar study in Indian eyes reported significantly thinner anterior sclera along the inferior meridian in high myopic eyes than emmetropic and mild to moderate myopic eyes.⁴ Most recently, Sung et al.¹⁵ also found similar trends in the distribution of AST in Koreans and reported a negative relationship between axial length and AST along temporal, nasal, and inferior meridians. Therefore, there occur consistent variations in AST between different ocular meridians, despite differences in ethnicity and degree of myopia in the study population. Together, the studies indicate that AST might act as an important biomarker for myopia progression.^4,15,22–24 

Various noncontact optical biometers based on the principle of optical low coherence reflectometry (OLCR), partial coherence interferometry (PCI), and swept-source OCT (SS-OCT) have been well accepted in ophthalmic practice.³² While these biometers are widely used primarily for intraocular lens (IOL) power calculation before cataract surgery,^32,33 they are also well-established diagnostic instruments in myopia management practice and research to accurately determine the biometric details of the eye, including axial length, which is a key determinant of axial myopia.^34–36 Previous investigations that attempted to compare various parameters recorded with a noncontact optical biometer and an OCT indicated a good correlation and agreement for axial length,³⁷ keratometry,³⁷ and anterior chamber depth measurements.^37,38 Furthermore, Read et al.³⁹ determined the retinal and choroidal thickness through a deliberate analysis of A-scan peaks obtained from the OLCR biometer (Lenstar LS 900; Haag Streit, Köniz, Switzerland) and reported a good agreement with a high-resolution spectral domain OCT (SD-OCT). 

The AST has so far been determined majorly with the OCT techniques.^4,15,22–27 While modern OCT instruments possess various advantages, they are relatively expensive with restricted portability and limited to a few researchers and practitioners. Considering a potential association between AST and myopia, as well as the ubiquitous use of optical biometers in clinical practice and research, we applied and validated a biometry-based technique for estimating AST based on the OCT landmarks. 

Method

This is a cross-sectional study where healthy participants without ocular or systematic pathologic conditions were recruited. Participants were the optometry students and staff members of the L V Prasad Eye Institute (LVPEI), Hyderabad, India. The study was approved by the institutional review board of LVPEI (LEC-BHR-P-07-21-723) and all the procedures adhered to the tenets of the Declaration of Helsinki. Appropriate written informed consent was obtained from all the participants included in the study. 

All participants (N = 62) aged 20 to 37 years underwent a comprehensive eye examination to determine normal ocular health and to rule out any pathology, inflammation, and degeneration of the anterior ocular surface. Noncycloplegic central refraction was measured in both eyes using an open-field autorefractor (WAM-5500; Grand Seiko, Hiroshima, Japan) under normal room illumination (200−250 lux). Refractive error was determined using the previously used standard protocol,^40–43 where the participant fixated at a high-contrast Maltese cross. This target was placed at the participant's line of sight at a distance of 3 m from the eye to minimize the active accommodative state. The average of five consecutive readings was considered the final refractive error. The participants were later divided into two refractive groups based on the spherical equivalent refraction (SER, defined as spherical power plus half of the cylindrical power): (1) nonmyopes (emmetropia to low hyperopia, SER: >−0.50 to ≤+1.00 D; n = 25) and (2) myopes (SER: ≤−0.50 to ≥−12.00 D; n = 37).⁴⁴ All participants had their monocular best-corrected distance and near visual acuity of 0.00 logarithm of the minimum angle of resolution (logMAR) or better when assessed with the high-contrast Bailey−Lovie LogMAR chart in normal room illumination. 

The AST was measured in the right eye using an SS-OCT (Topcon-3 DRI Triton Plus; Topcon Corporation, Tokyo, Japan) and an OLCR-based noncontact optical biometer (Lenstar LS 900, Haag-Streit, Köniz, Switzerland), whereas the fellow eye (left eye) was occluded to ensure appropriate fixation of the right eye. Participants were instructed to fixate on the external targets (Maltese cross) placed at ±30° eccentricity from the central fixation along the horizontal and vertical meridians to measure AST along all four meridians (superior, inferior, nasal, and temporal). Fixating to such an external target led the anterior sclera of the same eye in the antagonistic meridian to be exposed and aligned in front of the instrument. For instance, when the participant's right eye fixated on an external target at 30° in the nasal visual field, the same eye's temporal sclera was exposed (Fig. 1A). All the measurements were obtained during the same session on the same day by a single experienced examiner. However, to ensure unbiased measurement and documentation of AST determined with two instruments (OCT and biometer), image analysis and data extraction were conducted on 2 different days. 

Figure 1.

 

Determining the AST with SS-OCT. A 16-mm single-line scan passing through the temporal scleral reflex of the right eye when the participant fixated on a target located at 30° in the nasal visual field (A). Corresponding cropped raw B-scan image of the anterior sclera (B), where the downward-facing blue arrowheads indicate the location of episcleral blood vessels, which mark the anterior scleral margin, whereas the upward-facing yellow arrowheads mark the location of posterior scleral margin. A single downward-facing red arrowhead represents the location of the scleral spur (reference point on the limbal area). In the magnified image (C), the anterior scleral margin is indicated by the blue arrowheads and the posterior scleral margin by yellow arrowheads. The AST was determined as the linear distance (µm) between the anterior and posterior scleral margins (double-headed black arrow) at the same distance from the limbus/scleral spur at which the biometer obtained the corresponding measurement for each meridian. The purple and green diamonds represent the anterior conjunctival and episcleral thickness, respectively.

AST Determination Using OCT

For AST measurements with SS-OCT, similar protocols used in a previous study⁴ were followed, where an external anterior segment module was attached in front of the objective lens of the OCT instrument to capture images of an anterior sclera. The previous study also reported the intersession repeatability of AST to be 10 ± 12 µm.⁴ A 16-mm single-line scan protocol (length, 16 mm; depth, 2.6 mm) was used to capture the raw B-scan OCT images of the anterior sclera in each of the four ocular meridians (line [V] anterior segment and line [H] anterior segment for measurements along vertical and horizontal meridians, respectively). The line scan was placed such that it included the peripheral cornea and passed through the scleral reflex in each of the gaze positions (Fig. 1A). To ensure the highest possible imaging quality, the OCT scan was repeated if the signal strength of the obtained B-scan image was <50. 

The captured individual raw B-scan OCT images of the anterior sclera (dimensions: 1024 × 165 pixels) were manually analyzed using IMAGEnet 6 (version 1.22.1.14101) software. This involved identification of various tissue interfaces from anterior to posterior within the region of interest at the anterior sclera (Fig. 1B). In the magnified B-scan image (Fig. 1C), the anterior-most hyporeflective layer denotes the anterior conjunctival surface including tear film (purple diamond), followed by the second thin layer of hyperreflective region denoting the conjunctival−episcleral interface (green diamond) and immediate hyporeflective region denoting episcleral blood vessels, which also mark the anterior margin of anterior sclera (downward-facing blue arrowheads).⁴ Likewise, the posterior margin of the anterior sclera was identified as a demarcated line between the hyperreflective scleral tissue and hyporeflective ciliary body tissue (upward-facing yellow arrowheads). The scleral spur (a slightly depressed region in the limbal area, facing the anterior chamber) was identified as a reference point in the limbal area (downward-facing red arrowhead) to measure AST at a specified distance from the limbus. The anterior conjunctival and episcleral thicknesses were determined as the distance (µm) from the anterior conjunctival surface including tear film to the conjunctival−episcleral interface and conjunctival−episcleral interface to episcleral blood vessels, respectively. The AST with OCT was determined as the distance (µm) between the anterior and posterior scleral margins (Fig. 1C). It took approximately 2 to 3 minutes for AS-OCT to obtain B-scan images of the anterior sclera (one eye, right eye) along all the four meridians, and the time required to analyze the corresponding B-scan images to determine the anterior conjunctival, episcleral, and scleral thickness ranged between 4 and 5 minutes. To ensure both OCT and biometer obtained the measurements in the same region of the anterior sclera, the thickness measurements with OCT in each ocular meridian were determined at the same distance from the limbus at which the biometer obtained the corresponding measurements. 

AST Determination Using a Biometer

The Lenstar is used to determine the on-axis ocular biometric parameters. Ocular biometry measurements are typically performed by focusing the biometer graticule on the anterior cornea and aligning it with the two concentric reference circles/rings of 16 markers each while the participant is instructed to fixate an internal fixation (red light) target. A similar protocol was used to determine the AST with a biometer, where the graticule was focused on the anterior sclera and aligned with reflex obtained from the tear film on the conjunctival surface (Purkinje image in case of on-axis measures through the cornea) while the participant fixated on an external target in each meridian (Fig. 2A). An average of five such consecutive measurements was considered the final measurement at each of the measured ocular meridians. 

Figure 2.

 

Determining the AST with a noncontact optical biometer. A participant's right eye fixating an external Maltese cross target placed at 30° in the nasal visual field such that the temporal sclera was exposed and aligned with the biometer (A). The biometer graticule focused on the anterior sclera and aligned with the tear film reflex while the participant fixated on an external target. Determining the diameter of a measurement circle, so that the distance from the limbus at which the biometer obtained the measurements (radius of the graticule) can be computed (B). Illustration of the biometer's A-scan signal from the anterior sclera (C). Analysis of the anterior section of the A-scan signal (magnified view; D) to determine the interpeak distances allowing the measurement of anterior scleral tissue. The series of peaks (anterior to posterior) from the anterior section of the A-scan was known to correspond to the anterior bulbar conjunctiva, including the tear film (first peak, P1), anterior episclera (second peak, P2), anterior margin of the anterior sclera (third peak, P3), and posterior margin of the anterior sclera (fourth peak, P4). The AST was determined as the interpeak distance (µm) from the anterior (P3) to posterior margin (P4) of the anterior sclera.

To determine the location of the reflex that represented the region at which the measurements were recorded, distance from the limbus was computed manually by applying the inbuilt white-to-white measurement tool in the default Lenstar software (EyeSuite, version i9.6.3.0). This tool under the “Result Overview” section in the software provides access to the measurement circle. This circle was adjusted such that the reflex aligned with the central crosshair of the circle, and the circumference of the circle was aligned with the corresponding limbal region. The diameter (d) of each of the five measurement circles was determined (Fig. 2B). The radius of each circle was calculated [radius (r) = one-half of diameter (d)], which was determined to be the distance from the limbus at which the biometer obtained the measurement of anterior scleral tissue. This was particularly important to identify the specific location on the anterior sclera where the biometer obtained the measurements so that the corresponding OCT measures could also be determined on the same location on the anterior sclera. The distance from the limbus across all ocular meridians at which the biometer and OCT obtained the measurements ranged between 2.02 and 4.20 mm with the overall mean ± SD distance of 3.13 ± 0.88 mm. 

The A-scan biometry peaks obtained from the biometer (Fig. 2C) were analyzed adopting a previously used protocol by Read et al.,³⁹ who manually determined the retinal and choroidal thickness using an OLCR-based noncontact optical biometer (Lenstar). In the current study, this involved magnifying the anterior section of A-scan signals (by two steps) generated from conjunctival and scleral tissues, to identify the peaks representing these structures (Figs. 2C, 2D). Given that biometry was performed on the anterior scleral region, the four prominent A-scan peaks were expected to originate from structures like the conjunctiva, episclera, and sclera (Fig. 2D). The first prominent peak was considered to originate from the anterior bulbar conjunctiva including tear film (P1), the second prominent peak from the anterior episclera (P2), the third prominent peak from the anterior margin of the anterior sclera (P3), and fourth prominent peak from the posterior margin of the anterior sclera (P4). The first peak P1 was automatically detected by the Lenstar's EyeSuite software, whereas the in-built movable circular cursors/markers were aligned manually to determine the succeeding peaks, P2, P3, and P4. Finally, the last prominent peak, P4, after which the A-scan signal dampens was identified toward the end of the A-scan image. Performing optical biometry on anterior sclera (one eye, right eye) along all four meridians required about 5 to 6 minutes and postprocessing of the obtained corresponding A-scan images (five measurements) that involved determination of the distance from the limbus to the anterior scleral region where biometry was performed, and analysis of A-scan biometry peaks took approximately 10 minutes. 

Figure 3 shows the origin of A-scan peaks from corresponding anterior tissue interfaces. This way, the origin of each A-scan peak was validated using the OCT landmarks. The individual raw A-scan peaks originating from anterior ocular tissue that was obtained from six randomly chosen participants (right eyes, n = 6) are provided in Figure 4. The examiner was trained to identify these four prominent A-scan peaks through multiple manual analyses of A-scan images based on the OCT landmarks before starting a formal analysis for the study purpose. The conjunctival, episcleral, and scleral thicknesses determined from the biometer were consistent with the corresponding tissue thickness measurements from SS-OCT. 

Figure 3.

 

Representative image for the origin of biometer A-scan peaks with the corresponding SS-OCT B-scan anterior tissue interfaces.

Figure 4.

 

The individual raw A-scan peaks originating from anterior ocular tissue that was obtained from six randomly chosen participants (right eyes, n = 6) with a noncontact optical biometer (Lenstar LS-900). The four prominent A-scan peaks (P1, P2, P3, and P4) were identified, and anterior tissue thickness from the anterior region of the eye was determined using the protocol described in Figure 2D.

Statistical Analysis

SPSS Statistics, version 26.0 (IBM Corp., Armonk, NY, USA) was used to perform all the statistical analysis and the inbuilt features of Excel 2019 (Microsoft Corp., Redmond, WA, USA) were used to plot the graphs. The Kolmogorov–Smirnov test indicated that the overall AST data with both OCT and biometer were normally distributed across all the meridians, and thus the parametric tests were applied to test the statistical significance. Bland−Altman analysis, Pearson's correlation coefficient, and coefficient of variation were used to respectively determine the agreement, correlation, and variance between the AST measurements obtained with the OCT and biometer. Independent samples t-test was used to determine if the differences in tissue thickness were statistically significant between nonmyopes and myopes. A P value of <0.05 was considered statistically significant. 

To determine the intrasession repeatability for the measurements of anterior conjunctival, episcleral, and scleral thickness with the biometer, the five repeated measures within a single session performed on a subset of randomly chosen participants (n = 10) were analyzed using the repeated-measures analysis of variance test with Bonferroni post hoc correction for each of the four ocular meridians. Similarly, to determine the intersession repeatability of the optical biometer, the measurements between the two different sessions performed at the same time of the day on two different days on a subset of participants (n = 15) were analyzed using the paired samples t-test for each measurement meridians. For both the intra- and intersession repeatability of the biometer, the coefficient of variation was reported for anterior conjunctival, episcleral, and scleral thickness at each ocular meridian. 

Results

The mean ± standard deviation (SD) age of all participants (N = 62, 29 males) was 25 ± 3 years. The mean SERs of nonmyopes (n = 25) and myopes (n = 37) were +0.26 ± 0.39 D and −3.20 ± 2.84 D, respectively, and the corresponding axial lengths were 23.44 ± 0.66 mm and 24.39 ± 1.23 mm. 

AST

The measurements of AST between the OCT and biometer correlated for all meridians (r ≥ 0.70, overall r = 0.82, P < 0.01; Fig. 5A). The overall coefficient of variation for AST across different meridians with both instruments ranged between 9% and 12%. The mean difference in AST measurements between the OCT and biometer (OCT – biometer) was 3 ± 2.8 µm (limits of agreement, −89 to +83 µm) across all measurement meridians, which was not statistically significant between the two instruments (P = 0.23; Fig. 5B). 

Figure 5.

 

Scatterplot for the correlation (A) and Bland–Altman plot for the agreement (B) between the OCT and biometer for AST measurements (data pooled for all participants across all meridians). The solid and dashed black lines (A), respectively, indicate the trendline best fitted for the trend in correlation and the 1:1 ratio line for the 100% correlation (r = 1.00). Similarly, the solid and two dashed black lines (B), respectively, represent the mean difference (OCT – biometer) and upper and lower limits of agreement between the OCT and biometer for AST measurements.

The means ± standard errors (SEs) AST derived from the biometer for superior, inferior, nasal, and temporal meridians were 451 ± 7 µm, 562 ± 7 µm, 517 ± 7 µm, and 512 ± 7 µm, respectively, for all participants (Fig. 6A); corresponding values with the OCT instrument were 433 ± 6 µm, 578 ± 7 µm, 525 ± 8 µm, and 519 ± 7 µm. The mean AST was thickest in the inferior meridian and thinnest in the superior meridian with both instruments (Fig. 6A). When the AST was compared between the two refractive groups, the overall AST for all meridians with OCT was 520 ± 7 µm and 509 ± 6 µm, respectively, in nonmyopes and myopes (P = 0.27); corresponding overall ASTs with biometer were 513 ± 6 µm and 508 ± 5 µm (P = 0.58). On analyzing AST at different meridians between these two refractive groups, the pattern of AST at different meridians was similar (P ≥ 0.14 at each meridian) between nonmyopes (with OCT, superior: 436 ± 7 µm, inferior: 577 ± 10 µm, nasal: 534 ± 12 µm, and temporal: 531 ± 11 µm; with biometer, superior: 449 ± 9 µm, inferior: 567 ± 8 µm, nasal: 515 ± 10 µm, and temporal: 520 ± 11 µm) and myopes (with OCT, superior: 430 ± 9 µm, inferior: 579 ± 9 µm, nasal: 517 ± 9 µm, and temporal: 510 ± 8 µm; with biometer, superior: 452 ± 9 µm, inferior: 558 ± 11 µm, nasal: 517 ± 10 µm, and temporal: 505 ± 9 µm). The overall mean differences in AST measurements between SS-OCT and optical biometer (OCT – biometer) in any meridian ranged between −18 and +16 µm, and they were not significantly different between nonmyopes and myopes (P ≥ 0.17 for all ocular meridians; Fig. 6B). There was a greater thickness difference along the temporal meridian than other meridians between myopes (thinner) and nonmyopes, but it did not attain a statistical difference. 

Figure 6.

 

Mean ± standard error (SE) AST at different measurement meridians for all participants using a noncontact optical biometer (empty black circles with solid line) and SS-OCT (empty black triangles with dashed line) (A). An asterisk (*) indicates a statistically significant difference in the mean AST measurements between the superior and inferior vertical meridians with both OCT and biometer. Mean difference ± SE for the measurements of AST between the OCT and biometer (OCT – biometer ) along different meridians for all the participants (gray bars), nonmyopes (green bars), and myopes (blue bars) (B).

Given the previous evidence of inclusion of anterior episcleral tissue thickness for determining the AST,²² we also computed the total AST by adding the anterior episcleral thickness to the AST in the current study (total AST = anterior episcleral thickness + AST). The mean ± SE total AST with biometer was 612 ± 15 µm, 685 ± 14 µm, 659 ± 20 µm, and 641 ± 18 µm, respectively, for superior, inferior, nasal, and temporal meridians; corresponding values with OCT instrument were 589 ± 19 µm, 699 ± 9 µm, 659 ± 21 µm, and 635 ± 19 µm. The pattern of mean total AST was similar to and consistent with that of the AST pattern as reported in Figure 6A. The mean differences in total AST measurements between the OCT and biometer were also not statistically significant for all ocular meridians ( P ≥ 0.13). 

Anterior Conjunctival and Episcleral Thickness

The mean anterior conjunctival and episcleral thicknesses with the biometer and OCT instrument were found to be similar (Fig. 7). The mean differences (OCT – biometer) for the conjunctival and episcleral thickness ranged from −3 to −9 µm and +2 to −10 µm, respectively (P ≥ 0.12 for all ocular meridians). The mean conjunctival and episcleral thicknesses with both instruments were found to be thickest along temporal and superior meridians, respectively (Fig. 7). The overall coefficient of variation for anterior conjunctival and episcleral thickness across different meridians with both instruments ranged from 16% to 19% and 20% to 24%, respectively. 

Figure 7.

 

Mean ± SE conjunctival (solid lines) and episcleral (dashed lines) thickness at different measurement meridians using an optical biometer (empty black circles) and an SS-OCT (empty black triangles). An asterisk (*) along the solid and dashed lines, respectively, indicates the statistically significant differences in the mean conjunctival (nasal versus temporal) and episcleral (superior versus inferior) thickness measurements with both OCT and biometer.

Repeatability of the NonContact Optical Biometer

For intrasession repeatability of the optical biometer, the overall coefficients of variation for the measurements of anterior conjunctival, episcleral, and scleral thickness were 18.1%, 25.3%, and 10.5%, respectively, across all measurement meridians; the corresponding coefficients of variation for intersession repeatability of the biometer were 14.8%, 21.0%, and 12.5% (Table). In addition, when the intersession repeatability of the biometer was assessed on the same day with a gap of a few minutes in a subset of 15 participants, the coefficients of variation were 12.7%, 22.6%, and 9.6%, respectively for the measures of anterior conjunctival, episcleral, and scleral thickness. 

Table.

 

Intra- and Intersession Repeatability of the Noncontact Optical Biometer for the Anterior Conjunctival, Episcleral, and Scleral Thickness Measurements for all Ocular Meridians

 

Discussion

In the current study, we applied and validated the biometry-based technique for estimating AST by comparing it against a widely accepted and commercially available high-resolution SS-OCT. The results indicate a good agreement and strong correlation between the biometer and OCT for AST across all measured meridians. The manually derived measures of the abovementioned anterior tissue thickness with the biometer showed good intra- and intersession repeatability. 

We found similar trends in the mean anterior conjunctival, episcleral, and scleral thickness variations between different ocular meridians with both biometer and OCT. The mean AST was found to be thickest along the inferior meridian and thinnest along the superior meridian with both instruments, which is consistent with a previous report from our group that investigated the variation in AST along different ocular meridians using a high-resolution SS-OCT.⁴ Similar AST patterns across different meridians in both nonmyopes and myopes in our study may be due to the fact that most myopic participants were low myopes (SER >−6.00 D, n = 31) with only a few high myopes (SER ≤−6.00 D, n = 6). In the current study, the mean conjunctival and episcleral thicknesses with both instruments were found thickest along temporal and superior meridians, respectively. It is well established that the change in tissue interface during imaging with an OCT and a biometer is indicated by the change in reflectivity in an OCT's B-scan image and biometry's A-scan peaks at each tissue interface, respectively. Therefore, similar and consistent trends with both instruments for the measured tissue thickness validate the origin of the determined A-scan biometry peaks, that is, “P1,” “P2,” “P3,” and “P4” (Fig. 2D), based on OCT landmarks (Fig. 1C), which represent the anterior bulbar conjunctival surface, anterior conjunctival−episcleral interface (episclera), anterior margin of the anterior sclera, and posterior margin of the anterior sclera, respectively. 

Read et al.³⁹ determined the retinal and choroidal thickness via manual analysis of A-scan peaks obtained from the OLCR-based Lenstar biometer. A strong correlation was reported between the biometer and a high-resolution SD-OCT for the measurement of the foveal retinal thickness (r² = 0.80 and mean difference = 1 ± 8 µm) and subfoveal choroidal thickness (r² = 0.94 and mean difference = 4 ± 24 µm).³⁹ Using a similar protocol for analyzing A-scan peaks obtained from the anterior region of the eye in the current study, we found a good agreement (mean difference = 3 ± 2.8 µm and limits of agreement = −89 to +83 µm) and a strong correlation (r = 0.82) between biometer and SS-OCT for the measures of AST across all meridians. 

The coefficient of variation for AST with both instruments in this study ranged from 9% to 12% across all four meridians. The biometer demonstrated good intrasession repeatability, where the coefficient of variation for the measurements of AST ranged from 8.4% to 9.6% for all meridians; the corresponding coefficient of variation for intersession repeatability of the biometer was 7.9% to 13.3%. The interpretation of the coefficient of variation is reported to be excellent, good, acceptable, and poor when the measures of coefficient of variation are ≤10%, 10% to 20%, 20% to 30%, and >30%, respectively.⁴⁵ In the current study, the coefficient of variation for AST was ≤10% but <30% for conjunctival and episcleral thickness. These measures of the coefficient of variation for the tissue thickness at both intra- and intersession repeatability were similar to that of the coefficient of variation with both instruments for all meridians. Therefore, it is acceptable to indicate that the Lenstar biometer can also be used in addition to the SS-OCT instrument to determine tissue thickness from the anterior region of the eye. However, for research purposes, particularly in clinical trials, where using one instrument throughout the study duration is preferrable for a specific measurement, the interchangeable usage of these two instruments should be performed with caution. 

The modern noncontact optical biometers are widely used in ophthalmic practice for various presurgical screening as well as diagnosis and monitoring a wide range of ocular diseases and disorders.^32,33 The repeatability and reproducibility of the biometry data obtained with various commercially available noncontact optical biometers, including the one based on the SS-OCT (IOL Master 700) and PCI/OLCR (IOL Master 500/Lenstar) technique in healthy children, adults, and cataract patients, are excellent.^46–48 In addition, these optical biometers are also highly demanded to precisely measure the biometric details of the eye, including axial length, in the field of myopia research and practice.^34–36 A good correlation and agreement were reported between an SS-OCT and PCI-based biometer for axial length,³⁷ keratometry,³⁷ and anterior chamber depth measurements.^37,38 Whereas a few other studies that attempted to compare the various measures of ocular biometry between the OCT instrument and PCI- or OLCR-based optical biometer reported significant differences in axial length,⁵³ keratometry,⁵³ central corneal thickness,^38,53,54 anterior chamber depth,^53,55 and anterior chamber width measurements.³⁸ 

Buckhurst et al.²² reported thicker AST than in the current study. The previously used AS-OCT instrument had poor axial resolution (18 µm), which may have made it difficult to demarcate episclera from the sclera. Therefore, the thicker AST measurements in the previous study could be due to the inclusion of episcleral tissue thickness while determining the thickness of the anterior sclera. In the current study, when the total AST was computed by adding the anterior episcleral thickness to AST (total AST = anterior episcleral thickness + AST), the pattern of mean total AST was similar to and consistent with that of the mean AST pattern across all measurement meridians. The imaging instruments used in the current study have a high axial resolution (8 µm with SS-OCT and 10 µm with Lenstar biometer), which easily demarcates episclera from sclera, making an easy determination of AST in addition to the conjunctival and anterior episcleral thickness. For imaging tissue thickness, the noncontact biometer obtains biometric measures from only a single location at the anterior sclera (i.e., a one-dimensional scan), unlike two- and three-dimensional scans with a larger diameter with the modern SS-OCT instruments. However, a recent investigation reported that the high myopes demonstrate significant thinning of the anterior sclera inferiorly up to 3 mm from the limbus.⁴ In the current study, both the biometer and OCT obtained the measurements at a mean distance of 3.13 ± 0.88 mm (range, 2.02 to 4.20 mm) from the limbus along all ocular meridians. Therefore, it is reasonable to image the anterior sclera at around 3 mm from the limbus given the importance of the anterior sclera within this region, especially in myopes. Although the instruments used in the current study use a different refractive index for different ocular structures, the philosophy and mechanism based on which these instruments function do not seem to be a matter of concern given that the results from both instruments were similar. 

Major strengths of the current study include repurposing the usage and establishing the novelty of the Lenstar biometer to determine the AST in addition to ocular biometry. The limitation of this study involves the need for the examiner's rich experience and knowledge to identify the A-scan peaks to determine the tissue thickness. We also found that it was difficult to identify the peaks (especially P4) in approximately 6% to 8% of the participants, for whom the biometry was repeated to obtain the measurements of tissue thickness. While the direct export of data from the instrument, particularly the optical biometer, is preferable for research purposes and reducing postprocessing time (approximately 10 minutes in the current study), anterior scleral data exported from the currently available version of Lenstar EyeSuite software may not be accurate enough to detect the A-scan peaks (except P1) originating from the anterior scleral structures. Therefore, one experienced and well-trained examiner performed image acquisition and postprocessing manually in the current study. In future, updates in biometer software are required (which is outside the scope of the current study) to add “scleral mode” in addition to the currently available features that automatically detect the on-axis changes in tissue interfaces at the cornea, crystalline lens, and retina. Additionally, longitudinal studies based on machine learning models are required to identify the peaks and determine the interpeak distances more precisely. This will establish the practicality of the noncontact optical biometer in the field of myopia research and practice to detect the change in the AST in high and/or progressive myopes. Also, the applicability of a biometer to measure the anterior tissue thickness in elderly individuals with less complacent sclera and diseased ocular surfaces, like those receiving chronic intraocular pressure−lowering drops, pathology, inflammation, and degeneration of anterior ocular surface, should be determined to identify the possible challenges in these individuals and the measures to overcome them. 

In conclusion, the results of the current study indicate that the noncontact optical biometer (resolution: 10 µm), which is typically used to determine axial length, can also be used to determine anterior tissue thickness. This biometry-based novel technique (deliberate analysis of A-scan peaks) is capable of accurately, repeatably, and interchangeably estimating the anterior conjunctival, episcleral, and scleral thickness based on the OCT (resolution: 8 µm) landmarks. The findings indicate the wider application of high-resolution optical biometers in the field of myopia research and practice. 

Acknowledgments

The authors thank Jacinth J. Priscilla for her initial work on the project. 

This research was presented at the Annual Meeting of the Association for Research in Vision & Ophthalmology (ARVO) 2023, New Orleans, United States, on April 23–27, 2023. 

Supported by the partnership grant between the L V Prasad Eye Institute and the Department of Biotechnology, Government of India (BT/PR32404/MED/30/2136/2019) and by the Hyderabad Eye Research Foundation, L V Prasad Eye Institute, India. 

Disclosure: S.K. Gupta, None; R. Dhakal, None; P.K. Verkicharla, None