December 2014
Volume 55, Issue 12
Free
Glaucoma  |   December 2014
Comparison of Kang's and Littmann's Methods of Correction for Ocular Magnification in Circumpapillary Retinal Nerve Fiber Layer Measurement
Author Notes
  • Orthoptics and Visual Science, Department of Rehabilitation, School of Allied Health Sciences, Kitasato University, Kanagawa, Japan 
  • Correspondence: Kazunori Hirasawa, Orthoptics and Visual Science, Department of Rehabilitation, School of Allied Health Sciences, Kitasato University, 1-15-1 Kitasato, Minami-ku, Sagamihara, Kanagawa 252-0374, Japan; [email protected]
Investigative Ophthalmology & Visual Science December 2014, Vol.55, 8353-8358. doi:https://doi.org/10.1167/iovs.14-15720
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Kazunori Hirasawa, Nobuyuki Shoji, Yukako Yoshii, Shota Haraguchi; Comparison of Kang's and Littmann's Methods of Correction for Ocular Magnification in Circumpapillary Retinal Nerve Fiber Layer Measurement. Invest. Ophthalmol. Vis. Sci. 2014;55(12):8353-8358. https://doi.org/10.1167/iovs.14-15720.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: To assess and compare the accuracy of Kang's method for ocular magnification correction in circumpapillary retinal nerve fiber layer (cpRNFL) thickness measurement with that of the currently used Littmann's method.

Methods.: A total of 148 eyes of 148 healthy participants underwent cpRNFL measurement without correction and with correction by Littmann's method using spectral-domain optical coherence tomography. Correction of ocular magnification by Kang's method is based on the observed uncorrected cpRNFL thickness. The accuracy of Kang's method was compared with that of Littmann's method for magnification correction by using the Pearson product-moment correlation coefficient (r) and Bland–Altman analysis.

Results.: A total of 132 eyes were assessed. The two methods used to correct the global cpRNFL thickness were strongly correlated (r = 0.940, P < 0.01), and the mean difference between the two methods was −0.4 μm, with 95% limits of agreement (LoA) of ±6 μm without systemic bias (P > 0.05). Although the cpRNFL thicknesses corrected with the two methods at quadrant and 12-o′clock sectors showed a strong correlation (r > 0.731, P < 0.01), high LoAs ranging from ±13.3 to ±27.9 and systemic biases were observed at nasal and inferior sectors.

Conclusion.: No difference was found between Kang's and Littmann's methods for correction of ocular magnification in global cpRNFL thickness measurement. However, with regard to magnification correction for sectoral cpRNFL thickness, further scrutiny of Kang's method is warranted, even in healthy participants, because of the high variability between Littmann's and Kang's methods.

Introduction
Depending on the device used for optical coherence tomography (OCT), the thickness of the circumpapillary retinal nerve fiber layer (cpRNFL) is generally measured on a circular scan, approximately 3.4 mm in diameter, centered on the optic nerve head, based on the Gullstrand schematic eye (corneal radius = 7.70 mm; refraction = 0 diopter; axial length = 24.39 mm). However, the apparent size of the optic nerve head on the fundus photograph varies among individuals because the ocular magnification of individual optical systems is different.13 Therefore, when measuring the thickness of a specific area such as the cpRNFL, the size of the scan circle should be corrected according to the individual's optical system to obtain an accurate measurement. 
Traditionally, correction of ocular magnification in OCT is generally performed with Littmann's method,4,5 which corrects for the actual radius of scan circle on the disc by using the following formula:  where t is the actual fundus dimension, p is the magnification factor of the imaging system, q is the magnification factor for the individual eye, and s is the value obtained from the imaging device (i.e., cpRNFL thickness measured along a circle 3.4 mm in diameter). The ocular magnification factor q of the eye can be determined by the following formula: q = 0.01306 * (axial length − 1.82).6 Furthermore, p is a constant in a telecentric system.  
At present, few OCT devices are capable of correcting for ocular magnification with Littmann's method. Kang et al.7 proposed a new approach to correct for ocular magnification in global cpRNFL thickness measurement, which did not involve correcting for the actual radius of the scan circle on the disc using axial length, but instead, its calculations were based on the observed uncorrected cpRNFL thickness and axial length. Kang's method allowed researchers to correct the cpRNFL thickness using OCT devices that could not apply Littmann's method for the correction of ocular magnification813 because Kang's method can easily correct for ocular magnification in cpRNFL measurement if the axial length and uncorrected cpRNFL thickness were determined in advance. However, to our knowledge, no study has assessed the accuracy of the correction made by Kang's method. The aim of this study was to evaluate the accuracy of Kang's method for the correction of ocular magnification, which is based on the observed cpRNFL thickness, and to compare it with Littmann's method, which is based on the actual radius of the scan circle on the disc. 
Methods
A total of 148 student volunteers from Kitasato University were recruited for this prospective study. All the volunteers underwent comprehensive ophthalmic examinations, including noncycloplegic refraction (KR-8100PA; Topcon, Tokyo, Japan), visual acuity at 5 m by a Landolt ring chart, IOP (NT-530P; NIDEK, Aichi, Japan), axial length measurement (OA-1000; TOMEY, Aichi, Japan), and fundus examination by a glaucoma specialist (NS). The inclusion criteria were as follows: corrected visual acuity of 20/20 or better, IOP of 21 mm Hg or less, normal optic disc appearance, and no fundus disease. The study followed the tenets of the Declaration of Helsinki, and written informed consent was obtained from each participant after approval was received from the Ethics Committee of Kitasato University School of Allied Health Science (No. 2012–07). 
Optical Coherence Tomography Device
The cpRNFL was measured by a spectral-domain OCT (3D OCT-2000, version 8.00; Topcon). This device operates at a scan speed of 50,000 A-scans per second and has a depth and lateral resolution of 6 and 20 μm or less, respectively. It requires a pupil size of 2.5 mm or larger for imaging. The measurements were performed in three-dimensional optic disc scan mode, consisting of 512 A-scans per B-scan × 128 C-scan resolution in a 6 × 6-mm scan area, and cpRNFL thickness was analyzed by auto segmentation. In this mode, global and sectoral (quadrant and clock-hour) cpRNFL thickness measurements could be obtained from the images. Using 3D-OCT 2000, the cpRNFL thickness without magnification correction was measured in a 3.4-mm diameter scan circle centered on the optic nerve head, based on the Gullstrand schematic eye (corneal radius = 7.7 mm; refraction = 0 diopter; axial length = 24.39 mm). Although the detailed formula used for the correction of ocular magnification with this device has not been made public, it uses Littmann's method to correct for the scan circle diameter in the measurement of cpRNFL thickness by determining the refraction, corneal curvature, and axial length before imaging. Theoretically, correction for ocular magnification with Littmann's method can be performed after imaging if cpRNFL was imaged in the raster scan mode, because the analyzed scan circle diameter was corrected based on the formula. However, the current version of this device (version 8.00) is not built to correct for scan circle diameter after imaging. Thus, cpRNFL was measured both with and without Littmann's correction. 
Circumpapillary RNFL Measurement
Optical coherence tomography images of one randomly selected eye of each participant were obtained for evaluation of the cpRNFL, with and without correction by Littmann's method. Both images were captured by two examiners (YY, SH) without the use of a mydriatic agent. Both examiners evaluated an equal number of eyes (74 eyes each). The cpRNFL thickness was corrected by Kang's method, which was based on the axial length and observed uncorrected cpRNFL thickness, using Microsoft Excel (Microsoft Corporation, Redmond, WA), according to the method described in a previous study.7 The following formula was used:    
Furthermore, the cpRNFL thickness was assessed not only globally but also according to quadrants and clock-hour sectors. Mirror images of the right eye were used as results for the left eye. 
Next, the repeatability of the data also was assessed. Fifty-four eyes of 54 randomly selected participants were measured twice by both the examiners, with and without magnification correction, to investigate intraexaminer and interexaminer measurement errors. For the analysis of the interexaminer measurement error, the first measurement obtained by the two examiners was assessed. 
The exclusion criteria for analysis in this study were as follows: image quality of 70 or less and lack or deviation of B-scan line images. 
Statistical Analysis
All the data were compiled into a Microsoft Excel worksheet and analyzed by using statistical software (MedCalc version 13.2.0.0; MedCalc Software, Ostend, Belgium). Pearson product-moment correlation coefficient (r) and Bland–Altman analysis were used to compare the cpRNFL thicknesses corrected by the Littman's and Kang's methods.14 Using the Bland–Altman plot to assess the fixed bias, we evaluated whether the 95% coefficient interval of the mean difference between Littman's and Kang's methods included “zero.”14 The following formula was used:  where is the mean difference of cpRNFL thickness corrected by both the methods, t is the t distribution (n − 1, 0.05) of the degree of freedom (n − 1), n is the number of the eyes, and sd is the SD of the mean difference of cpRNFL thickness corrected by both methods.  
The proportional bias, with the Bland–Altman plot, was assessed on the basis of a P value of less than 5% in linear regression analysis. Limits of agreement (LoA) of the sample mean were calculated using the mean difference and its SD (sd).14 The following formula was used:     
In the current study, LoA with Bland–Altman analysis of the population mean was also based on the SE of LoA in a sample mean.14 The following formula was applied:     
The repeatability of intraexaminer and interexaminer measurements was assessed using the methods described above. 
Results
Of the initial 148 participants, 16 (6 and 10 participants measured by examiner YY and SH, respectively) were excluded, and, therefore, 132 eyes of 132 participants (68 and 64 participants measured by examiner YY and SH, respectively) were included in the final analyses. The demographic data of the study participants are shown in Table 1
Table 1
 
Demographic and Ocular Characteristics of the Participants
Table 1
 
Demographic and Ocular Characteristics of the Participants
Parameter Mean ± SD Range, Minimum–Maximum
Age, y 21.7 ± 1.6  20–28
Spherical, diopter −3.02 ± 3.08 −13.00–(+4.00)
Astigmatism, diopter −0.68 ± 0.83 −6.00–0.00
Average corneal radius, mm 7.79 ± 0.27 6.93–8.42
Visual acuity, logMAR −0.23 ± 0.07 −0.30–0.00
Axial length, mm 24.78 ± 1.51 21.25–28.35
Central corneal thickness, μm 539.3 ± 24.5 472.0–595.0
IOP, mm Hg 14.4 ± 2.2  7.7–20.0
The repeatability of the data obtained in the current study is shown in Table 2. The repeatability was found to be the worst when sector-wise assessment of the cpRNFL was performed. 
Table 2
 
Limits of Agreement for Intra- and Interexaminer Measurement Error of cpRNFL Thickness
Table 2
 
Limits of Agreement for Intra- and Interexaminer Measurement Error of cpRNFL Thickness
Sector Intraexaminer Interexaminer
Examiner YY Examiner SH Examiner YY-SH
Sample Population Sample Population Sample Population
Global ±4.5/±4.7 ±3.0/±3.1 ±3.8/±4.3 ±2.7/±2.9 ±5.2/±4.9 ±3.1/±3.0
Superior ±7.4/±7.8 ±4.2/±4.4 ±8.9/±9.6 ±4.9/±5.0 ±8.5/±8.6 ±4.1/±4.1
Nasal ±8.5/±9.2 ±4.6/±4.8 ±9.0/±8.5 ±4.9/±4.7 ±9.9/±10.8 ±4.4/±4.5
Inferior ±8.7/±8.6 ±4.7/±4.7 ±8.0/±9.2 ±4.5/±4.9 ±7.5/±9.8 ±3.9/±4.4
Temporal ±5.6/±6.8 ±3.5/±4.0 ±6.2/±9.4 ±3.8/±5.0 ±5.9/±8.3 ±3.4/±4.1
12 o′clock ±9.9/±11.6 ±5.0/±5.3 ±11.6/±11.3 ±5.6/±5.5 ±11.1/±12.1 ±4.5/±4.6
1 o′clock ±11.0/±10.7 ±5.3/±5.2 ±9.0/±10.3 ±4.9/±5.3 ±12.0/±9.5 ±4.6/±4.3
2 o′clock ±12.6/±9.7 ±5.6/±4.9 ±12.0/±11.4 ±5.7/±5.5 ±13.8/±13.4 ±4.7/±4.7
3 o′clock ±10.9/±10.7 ±5.2/±5.2 ±10.3/±9.8 ±5.3/±5.1 ±12.9/±13.3 ±4.7/±4.7
4 o′clock ±11.2/±13.1 ±5.3/±5.6 ±10.1/±9.9 ±5.2/±5.2 ±11.9/±10.0 ±4.6/±4.4
5 o′clock ±12.1/±12.0 ±5.5/±5.5 ±9.8/±9.9 ±5.1/±5.1 ±7.8/±11.0 ±3.9/±4.5
6 o′clock ±13.1/±12.2 ±5.6/±5.5 ±13.3/±13.2 ±5.9/±5.9 ±12.3/±14.0 ±4.6/±4.7
7 o′clock ±15.7/±14.0 ±5.9/±5.8 ±12.5/±12.4 ±5.7/±5.7 ±14.2/±13.7 ±4.7/±4.7
8 o′clock ±8.6/±8.4 ±4.6/±4.6 ±7.3/±9.7 ±4.3/±5.1 ±9.9/±9.2 ±4.4/±4.2
9 o′clock ±6.9/±7.4 ±4.1/±4.2 ±5.7/±7.0 ±3.6/±4.2 ±4.9/±8.6 ±3.0/±4.1
10 o′clock ±9.2/±8.7 ±4.8/±4.7 ±10.6/±8.6 ±5.3/±4.8 ±8.2/±10.0 ±4.0/±4.4
11 o′clock ±13.8/±12.5 ±5.7/±5.5 ±12.9/±14.0 ±5.8/±6.0 ±12.7/±14.4 ±4.7/±4.7
Table 3 shows the results of Pearson correlation coefficient and Bland–Altman analyses of cpRNFL thickness corrected by Littmann's and Kang's methods. The corrected cpRNFL thickness at each sector was strongly correlated between the two methods (r = 0.731–0.940 with P < 0.01). However, a 95% LoA increase was observed when the cpRNFL was assessed according to sectors. Fixed bias was observed at the quadratic nasal sector, as well as 3, 4, and 5 o′clock sectors. Proportional bias also was observed at the 6 o′clock sector (P < 0.01). The number of eyes for which the difference in the cpRNFL thickness corrected by Littmann's and Kang's methods was within the measurement error is shown in Table 4
Table 3
 
Evaluation of cpRNFL Thickness Corrected With Littmann's and Kang's Methods
Table 3
 
Evaluation of cpRNFL Thickness Corrected With Littmann's and Kang's Methods
Sector Pearson r Bland–Altman Analysis
95% LoA Fixed Bias Proportional Bias
Sample Population Mean Difference (95% CI) Slope P Value
Global 0.940** ±6.0 ±4.4 −0.4 (−1.0 to 0.1) 0.005 0.87
Superior 0.882** ±14.3 ±8.4 −0.4 (−1.6 to 0.9) 0.016 0.72
Nasal 0.903** ±13.3 ±8.0 −1.2 (−2.4 to −0.1)* 0.045 0.26
Inferior 0.851** ±14.9 ±8.7 −0.2 (−1.5 to 1.2) −0.099 0.07
Temporal 0.880** ±16.8 ±9.3 0.1 (−1.4 to 1.5) 0.000 0.99
12 o′clock 0.895** ±18.8 ±9.9 −0.5 (−2.2 to 1.1) −0.010 0.79
1 o′clock 0.873** ±19.0 ±10.0 −0.7 (−2.4 to 1.0) 0.037 0.42
2 o′clock 0.885** ±16.1 ±9.1 −1.4 (−2.8 to 0.0) −0.016 0.72
3 o′clock 0.860** ±15.1 ±8.7 −1.8 (−3.1 to −0.5)* 0.071 0.08
4 o′clock 0.902** ±13.6 ±8.2 −1.6 (−2.8 to −0.4)* 0.055 0.17
5 o′clock 0.910** ±15.1 ±8.7 −1.5 (−2.8 to −0.1)* 0.000 0.99
6 o′clock 0.876** ±20.9 ±10.5 1.0 (−0.9 to 2.8) −0.134* <0.01
7 o′clock 0.825** ±27.5 ±11.8 0.4 (−2.0 to 2.8) −0.083 0.13
8 o′clock 0.851** ±23.7 ±11.2 0.4 (−1.7 to 2.5) 0.015 0.77
9 o′clock 0.902** ±12.6 ±7.8 0.4 (−0.7 to 1.6) 0.053 0.18
10 o′clock 0.913** ±18.6 ±9.9 0.0 (−1.6 to 1.6) −0.056 0.14
11 o′clock 0.731** ±27.9 ±11.9 0.0 (−2.5 to 2.4) 0.014 0.84
Table 4
 
Number of Eyes for Which the Difference in the Corrected cpRNFL Thickness by Littmann's and Kang's Methods Was Within the Measurement Error
Table 4
 
Number of Eyes for Which the Difference in the Corrected cpRNFL Thickness by Littmann's and Kang's Methods Was Within the Measurement Error
Sector Maximum Value of 95% LoA, μm No. of Eyes (%)
Global 5.2 121 (91.7)
Superior 9.6 119 (90.2)
Nasal 10.8 119 (90.2)
Inferior 9.8 115 (87.1)
Temporal 9.4 123 (93.2)
12 o′clock 12.1 113 (85.6)
1 o′clock 12.0 117 (88.6)
2 o′clock 13.8 118 (89.4)
3 o′clock 13.3 119 (90.2)
4 o′clock 13.1 124 (93.9)
5 o′clock 12.1 115 (87.1)
6 o′clock 13.1 110 (83.3)
7 o′clock 15.7 112 (84.8)
8 o′clock 9.9 121 (91.7)
9 o′clock 8.6 117 (88.6)
10 o′clock 10.6 119 (90.2)
11 o′clock 14.4 111 (84.1)
Figure 1 shows scatter plots and Bland–Altman plots of the global cpRNFL thickness corrected by Littmann's and Kang's method. The global cpRNFL thickness corrected by both the methods was strongly correlated (r = 0.940, y = 7.2166 + 0.9358x, P < 0.01). According to Bland–Altman analysis, the cpRNFL thickness corrected by Littmann's and Kang's methods showed a mean difference of −0.4 μm with a 95% LoA of −6.5 to 5.6 μm of sample mean and −4.8 to 4.0 μm of population mean without fixed bias; moreover, no proportional bias was observed (y = −0.952 + 0.005x, P = 0.87). 
Figure 1
 
Scatter plots and Bland–Altman plots of the global cpRNFL thickness corrected with Littmann's and Kang's methods. Scatter plots (left) show the regression line (solid line) and its 95% confidence intervals (dashed line). The Bland–Altman plot (right) shows the mean difference between the two methods (solid line) and LoA (dashed line).
Figure 1
 
Scatter plots and Bland–Altman plots of the global cpRNFL thickness corrected with Littmann's and Kang's methods. Scatter plots (left) show the regression line (solid line) and its 95% confidence intervals (dashed line). The Bland–Altman plot (right) shows the mean difference between the two methods (solid line) and LoA (dashed line).
Discussion
In this study, we assessed the accuracy of Kang's method of correction for ocular magnification in cpRNFL thickness measurement and compared it with the currently used Littmann's method of correction. The highest correlation coefficient and best 95% LoA of the cpRNFL thickness corrected with Littmann's and Kang's methods were observed in measurement of the global cpRNFL thickness. However, these statistics gradually deteriorated when the cpRNFL was divided into quadrants and o′clock sectors. Fixed bias and proportional bias were observed at the quadratic nasal sector, as well as 3, 4, 5, and 6 o′clock sectors. 
The presence of systemic biases at the nasal region in the quadratic and o′clock sectors could be attributed to the magnitude of curvature of the fundus, because the scan circle is centered on the optic disc. In Littmann's method, the cpRNFL thickness was measured by the corrected scan circle size based on the individual optical system, whereas Kang's method linearly estimated the corrected cpRNFL based on the individual optical system and observed uncorrected cpRNFL thickness. Therefore, with Kang's method, the cpRNFL at the nasal region, where the magnitude of curvature is larger than that of the temporal region, would be overestimated, compared with Littmann's method, especially considering the longer axial length of a myopic eye. This is further supported by the results of this study: the mean difference of cpRNFL thickness at the nasal region, corrected with both methods, leans slightly to the negative side in Bland–Altman analysis (Table 3). 
In the current study, the differences in cpRNFL thicknesses corrected by Littmann's and Kang's methods were larger when the cpRNFL was divided into quadrants and o′clock sectors. This finding might be attributed to the fact that the same baseline cpRNFL thickness was not used to analyze the correction in both methods. The cpRNFL thickness with Kang's method was corrected based on the uncorrected baseline cpRNFL thickness, whereas Littmann's method (with the 3D-OCT 2000) required another imaging procedure for correcting the scan circle because the current version of this device cannot correct the scan circle after imaging. Because the measurement error gradually deteriorated when the cpRNFL was divided into small sectors (Table 2), the differences in cpRNFL thicknesses corrected by Littmann's and Kang's methods would be noted at o′clock sectors. To compare the corrected cpRNFL yielded by the two methods, it would be ideal to use the same baseline cpRNFL thickness. 
The intra- and interexaminer measurement errors of global cpRNFL thickness with and without correction ranged from 2.7 to 5.2 μm. The measurement errors of spectral-domain OCT are reportedly in the range of 3.1 to 11.7 μm.1528 The lower values in the present study can be attributed to good fixation during imaging because the participants were young. They also suggest lack of operator bias with regard to the imaging technique. 
In the current study, we also calculated the number of participants for whom the cpRNFL thickness at each sector, corrected by Littmann's and Kang's methods, was within the measurement error. Despite the difference between the two methods in the corrected cpRNFL sectoral thickness, the cpRNFL thickness at the global region was within the measurement error in 91.7% of the patients. Although the 95% LoAs of both the methods were slightly above the intraexaminer and interexaminer repeatability, the difference was within 1 μm in a sample mean of this study. Additionally, no systemic bias existed, and the highest correlation coefficient was determined at the global cpRNFL thickness. Considering these results, we speculated that Kang's method would be acceptable for correcting the global cpRNFL thickness. However, correction of sector-wise thickness with Kang's method would be challenging because of low repeatability compared with the global region, a finding similar to those reported in previous studies,1623,2528 the effects of ocular rotation on cpRNFL imaging,29 and the effect of temporal shift of the cpRNFL bundle caused by myopic changes.3032 
A major characteristic of Kang's method is that it can be used to correct the cpRNFL thickness, and not the actual radius of scan circle on the disc, using axial length and the observed cpRNFL thickness. Kang's method is based on the assumption that the same numbers of retinal nerve fibers cross the 1.73-mm radius scan circle.7 Furthermore, it was believed that the actual cpRNFL thickness at a 1.73-mm radius can be estimated from the cpRNFL thickness observed at a 1.73-mm radius before magnification correction. However, Kang's method cannot accurately estimate cpRNFL thickness with linear regression, especially at the regions nearer to the optic disc head, because cpRNFL near the optic disc is radially distributed. Varma et al.33 and Hirasawa et al.34 reported the distribution of cpRNFL thickness in human eyes by histologic and OCT analysis, respectively. Hirasawa et al.34 reported the cpRNFL thickness measured at 0.3-mm intervals from 2.2-mm to 4.0-mm radius of the scan circle centered on the optic disc. According to their study, the cpRNFL thickness increases linearly at radii of 3.1 to 4.0 mm of the scan circle, but the variation in cpRNFL thickness is not linear at radii of 2.2 to 2.8 mm of the scan circle. When the scan circle is corrected by using Littman's method, it would not affect the cpRNFL thickness. However, Kang's method, which corrects for the observed cpRNFL thickness, has the potential to overestimate the cpRNFL thickness. Therefore, further studies must be performed to assess the use of Kang's method of correcting for ocular magnification in cases of increased axial length. 
This study has the following limitations. First, although p value, which is the magnification factor of the imaging system, was a constant for the OCT device, it might vary among the different OCT devices. However, the p value of 3.382 was calculated with Stratus OCT or Cirrus HD OCT, and was quoted in this study based on previous reports,47,35 because the p value of 3D-OCT 2000 used in current study is also unknown to researchers. Therefore, the cpRNFL thickness corrected with Kang's method might be slightly different from that corrected with Littmann's method. However, it would have had an insignificant effect on cpRNFL thickness because the differences in p of the magnification factor of the imaging system would be quite negligible. Although the telecentric design with 3D-OCT 2000 has not been made public, we can assume it is because the default axial length, corneal radius, and refraction are determined as 24.39 mm, 7.7 mm, and 0 diopter, respectively, based on the Gullstrand schematic eye in 3D-OCT 2000. Littmann's method corrected for the ocular magnification by using Equation 1: t = p * q * s. In 3D OCT-2000, the magnification factor in an eye with an axial length of 24.38 mm is 1 (i.e., t = s). Substituting t = s into the formula, we arrive at p = 1/q and p can be calculated as 1/(0.01306 [24.38 − 1.82]) = 3.394. Additionally, because the corneal radius is considered in 3D-OCT 2000, both p values would be slightly different. However, a previous study reported that the effect of the corneal radius on magnification factor is very small.6 On the other hand, with Stratus OCT, the default axial length is 24.46 and refraction is 0 diopter.35 The Cirrus HD OCT system has the same magnification factor as that of the Stratus OCT system.7 Therefore, p can be calculated as 1/(0.01306 [24.48 − 1.82]) = 3.382 in Cirrus HD OCT. It can be speculated that the value of p obtained from Cirrus HD OCT and 3D OCT-2000 are nearly equal. Second, this study included only young healthy participants, and further studies in patients with glaucoma may be necessary. 
In conclusion, cpRNFL thickness corrected with Kang's method was as accurate as Littmann's method in correcting for ocular magnification at a global level in healthy young participants. The correction of cpRNFL thickness at each sector by Kang's method would require further assessment because of the high variability between Littmann's method and Kang's method. Studies involving patients with glaucoma and sector-wise correction with Kang's method should be performed in the future. 
Acknowledgments
Disclosure: K. Hirasawa, None; N. Shoji, None; Y. Yoshii, None; S. Haraguchi, None 
References
Coleman AL Haller JA Quigley HA. Determination of the real size of fundus objects from fundus photographs. J Glaucoma. 1996; 5: 433–435. [PubMed]
Dawczynski J Gebhardt M Mohr T Determination of absolute size of fundus objects. Graefes Arch Clin Exp Ophthalmol. 2011; 249: 381–387. [CrossRef] [PubMed]
Garway-Heath DF Rudnicka AR Lowe T Foster PJ Fitzke FW Hitchings RA. Measurement of optic disc size: equivalence of methods to correct for ocular magnification. Br J Ophthalmol. 1998; 82: 643–649. [CrossRef] [PubMed]
Littmann H. Determination of the real size of an object on the fundus of the living eye. Klin Monbl Augenheilkd. 1982; 180: 286–289. [CrossRef] [PubMed]
Littmann H. Determination of the true size of an object on the fundus of the living eye. By H. Littmann from the original article, “ Zur Bestimmung der wahren Grosse eines Objektes auf dem Hintergrund des lebenden Auges,” which originally appeared in Klinisches Monatsblatter fur Augenheilkunde 1982; 180: 286–9. Translated by TD Williams. Optom Vis Sci. 1992; 69: 717–720.
Bennett AG Rudnicka AR Edgar DF. Improvements on Littmann's method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol. 1994; 232: 361–367. [CrossRef] [PubMed]
Kang SH Hong SW Im SK Lee SH Ahn MD. Effect of myopia on the thickness of the retinal nerve fiber layer measured by Cirrus HD optical coherence tomography. Invest Ophthalmol Vis Sci. 2010; 51: 4075–4083. [CrossRef] [PubMed]
Savini G Barboni P Parisi V Carbonelli M. The influence of axial length on retinal nerve fibre layer thickness and optic-disc size measurements by spectral-domain OCT. Br J Ophthalmol. 2012; 96: 57–61. [CrossRef] [PubMed]
Aykut V Oner V Tas M Iscan Y Agachan A. Influence of axial length on peripapillary retinal nerve fiber layer thickness in children: a study by RTVue spectral-domain optical coherence tomography. Curr Eye Res. 2013; 38: 1241–1247. [CrossRef] [PubMed]
Oner V Aykut V Tas M Alakus MF Iscan Y. Effect of refractive status on peripapillary retinal nerve fibre layer thickness: a study by RTVue spectral domain optical coherence tomography. Br J Ophthalmol. 2013; 97: 75–79. [CrossRef] [PubMed]
Oner V Tas M Turkcu FM Alakus MF Iscan Y Yazici AT. Evaluation of peripapillary retinal nerve fiber layer thickness of myopic and hyperopic patients: a controlled study by Stratus optical coherence tomography. Curr Eye Res. 2013; 38: 102–107. [CrossRef] [PubMed]
Hsu SY Chang MS Ko ML Harnod T. Retinal nerve fibre layer thickness and optic nerve head size measured in high myopes by optical coherence tomography. Clin Exp Optom. 2013; 96: 373–378. [CrossRef] [PubMed]
Oner V Ozgur G Turkyilmaz K Sekeryapan B Durmus M. Effect of axial length on retinal nerve fiber layer thickness in children. Eur J Ophthalmol. 2014; 24: 265–272. [CrossRef] [PubMed]
Bland JM Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1: 307–310. [CrossRef] [PubMed]
Menke MN Knecht P Sturm V Dabov S Funk J. Reproducibility of nerve fiber layer thickness measurements using 3D Fourier-domain OCT. Invest Ophthalmol Vis Sci. 2008; 49: 5386–5391. [CrossRef] [PubMed]
Vizzeri G Weinreb RN Gonzalez-Garcia AO Agreement between spectral-domain and time-domain OCT for measuring RNFL thickness. Br J Ophthalmol. 2009; 93: 775–781. [CrossRef] [PubMed]
Gonzalez-Garcia AO Vizzeri G Bowd C Medeiros FA Zangwill LM Weinreb RN. Reproducibility of RTVue retinal nerve fiber layer thickness and optic disc measurements and agreement with Stratus optical coherence tomography measurements. Am J Ophthalmol. 2009; 147: 1067–1074, 1074.e1. [CrossRef] [PubMed]
Garas A Vargha P Hollo G. Reproducibility of retinal nerve fiber layer and macular thickness measurement with the RTVue-100 optical coherence tomograph. Ophthalmology. 2010; 117: 738–746. [CrossRef] [PubMed]
Mwanza JC Chang RT Budenz DL Reproducibility of peripapillary retinal nerve fiber layer thickness and optic nerve head parameters measured with cirrus HD-OCT in glaucomatous eyes. Invest Ophthalmol Vis Sci. 2010; 51: 5724–5730. [CrossRef] [PubMed]
Hong S Kim CY Lee WS Seong GJ. Reproducibility of peripapillary retinal nerve fiber layer thickness with spectral domain cirrus high-definition optical coherence tomography in normal eyes. Jpn J Ophthalmol. 2010; 54: 43–47. [CrossRef] [PubMed]
Seibold LK Mandava N Kahook MY. Comparison of retinal nerve fiber layer thickness in normal eyes using time-domain and spectral-domain optical coherence tomography. Am J Ophthalmol. 2010; 150: 807–814. [CrossRef] [PubMed]
Wu H de Boer JF Chen TC. Reproducibility of retinal nerve fiber layer thickness measurements using spectral domain optical coherence tomography. J Glaucoma. 2011; 20: 470–476. [CrossRef] [PubMed]
Shpak AA Sevostyanova MK Ogorodnikova SN Shormaz IN. Comparison of measurement error of Cirrus HD-OCT and Heidelberg Retina Tomograph 3 in patients with early glaucomatous visual field defect. Graefes Arch Clin Exp Ophthalmol. 2012; 250: 271–277. [CrossRef] [PubMed]
Toteberg-Harms M Sturm V Knecht PB Funk J Menke MN. Repeatability of nerve fiber layer thickness measurements in patients with glaucoma and without glaucoma using spectral-domain and time-domain OCT. Graefes Arch Clin Exp Ophthalmol. 2012; 250: 279–287. [CrossRef] [PubMed]
Tan BB Natividad M Chua KC Yip LW. Comparison of retinal nerve fiber layer measurement between 2 spectral domain OCT instruments. J Glaucoma. 2012; 21: 266–273. [CrossRef] [PubMed]
Carpineto P Nubile M Agnifili L Reproducibility and repeatability of Cirrus HD-OCT peripapillary retinal nerve fibre layer thickness measurements in young normal subjects. Ophthalmologica. 2012; 227: 139–145. [CrossRef] [PubMed]
Hong JT Sung KR Cho JW Yun SC Kang SY Kook MS. Retinal nerve fiber layer measurement variability with spectral domain optical coherence tomography. Korean J Ophthalmol. 2012; 26: 32–38. [CrossRef] [PubMed]
Moreno-Montanes J Olmo N Garcia N Alvarez A Garcia-Granero M. Influence of examiner experience on the reproducibility of retinal nerve fiber thickness values using Cirrus and Stratus OCTs. J Glaucoma. 2013; 22: 243–249. [CrossRef] [PubMed]
Kanamori A Nakamura M Tabuchi K Yamada Y Negi A. Effects of ocular rotation on parapapillary retinal nerve fiber layer thickness analysis measured with spectral-domain optical coherence tomography. Jpn J Ophthalmol. 2012; 56: 354–361. [CrossRef] [PubMed]
Kim MJ Lee EJ Kim TW. Peripapillary retinal nerve fiber layer thickness profile in subjects with myopia measured using the Stratus optical coherence tomography. Br J Ophthalmol. 2010; 94: 115–120. [CrossRef] [PubMed]
Yoo YC Lee CM Park JH. Changes in peripapillary retinal nerve fiber layer distribution by axial length. Optom Vis Sci. 2012; 89: 4–11. [CrossRef] [PubMed]
Yamashita T Asaoka R Tanaka M Relationship between position of peak retinal nerve fiber layer thickness and retinal arteries on sectoral retinal nerve fiber layer thickness. Invest Ophthalmol Vis Sci. 2013; 54: 5481–5488. [CrossRef] [PubMed]
Varma R Skaf M Barron E. Retinal nerve fiber layer thickness in normal human eyes. Ophthalmology. 1996; 103: 2114–2119. [CrossRef] [PubMed]
Hirasawa H Tomidokoro A Araie M Peripapillary retinal nerve fiber layer thickness determined by spectral-domain optical coherence tomography in ophthalmologically normal eyes. Arch Ophthalmol. 2010; 128: 1420–1426. [CrossRef] [PubMed]
Leung CK Cheng AC Chong KK Optic disc measurements in myopia with optical coherence tomography and confocal scanning laser ophthalmoscopy. Invest Ophthalmol Vis Sci. 2007; 48: 3178–3183. [CrossRef] [PubMed]
Figure 1
 
Scatter plots and Bland–Altman plots of the global cpRNFL thickness corrected with Littmann's and Kang's methods. Scatter plots (left) show the regression line (solid line) and its 95% confidence intervals (dashed line). The Bland–Altman plot (right) shows the mean difference between the two methods (solid line) and LoA (dashed line).
Figure 1
 
Scatter plots and Bland–Altman plots of the global cpRNFL thickness corrected with Littmann's and Kang's methods. Scatter plots (left) show the regression line (solid line) and its 95% confidence intervals (dashed line). The Bland–Altman plot (right) shows the mean difference between the two methods (solid line) and LoA (dashed line).
Table 1
 
Demographic and Ocular Characteristics of the Participants
Table 1
 
Demographic and Ocular Characteristics of the Participants
Parameter Mean ± SD Range, Minimum–Maximum
Age, y 21.7 ± 1.6  20–28
Spherical, diopter −3.02 ± 3.08 −13.00–(+4.00)
Astigmatism, diopter −0.68 ± 0.83 −6.00–0.00
Average corneal radius, mm 7.79 ± 0.27 6.93–8.42
Visual acuity, logMAR −0.23 ± 0.07 −0.30–0.00
Axial length, mm 24.78 ± 1.51 21.25–28.35
Central corneal thickness, μm 539.3 ± 24.5 472.0–595.0
IOP, mm Hg 14.4 ± 2.2  7.7–20.0
Table 2
 
Limits of Agreement for Intra- and Interexaminer Measurement Error of cpRNFL Thickness
Table 2
 
Limits of Agreement for Intra- and Interexaminer Measurement Error of cpRNFL Thickness
Sector Intraexaminer Interexaminer
Examiner YY Examiner SH Examiner YY-SH
Sample Population Sample Population Sample Population
Global ±4.5/±4.7 ±3.0/±3.1 ±3.8/±4.3 ±2.7/±2.9 ±5.2/±4.9 ±3.1/±3.0
Superior ±7.4/±7.8 ±4.2/±4.4 ±8.9/±9.6 ±4.9/±5.0 ±8.5/±8.6 ±4.1/±4.1
Nasal ±8.5/±9.2 ±4.6/±4.8 ±9.0/±8.5 ±4.9/±4.7 ±9.9/±10.8 ±4.4/±4.5
Inferior ±8.7/±8.6 ±4.7/±4.7 ±8.0/±9.2 ±4.5/±4.9 ±7.5/±9.8 ±3.9/±4.4
Temporal ±5.6/±6.8 ±3.5/±4.0 ±6.2/±9.4 ±3.8/±5.0 ±5.9/±8.3 ±3.4/±4.1
12 o′clock ±9.9/±11.6 ±5.0/±5.3 ±11.6/±11.3 ±5.6/±5.5 ±11.1/±12.1 ±4.5/±4.6
1 o′clock ±11.0/±10.7 ±5.3/±5.2 ±9.0/±10.3 ±4.9/±5.3 ±12.0/±9.5 ±4.6/±4.3
2 o′clock ±12.6/±9.7 ±5.6/±4.9 ±12.0/±11.4 ±5.7/±5.5 ±13.8/±13.4 ±4.7/±4.7
3 o′clock ±10.9/±10.7 ±5.2/±5.2 ±10.3/±9.8 ±5.3/±5.1 ±12.9/±13.3 ±4.7/±4.7
4 o′clock ±11.2/±13.1 ±5.3/±5.6 ±10.1/±9.9 ±5.2/±5.2 ±11.9/±10.0 ±4.6/±4.4
5 o′clock ±12.1/±12.0 ±5.5/±5.5 ±9.8/±9.9 ±5.1/±5.1 ±7.8/±11.0 ±3.9/±4.5
6 o′clock ±13.1/±12.2 ±5.6/±5.5 ±13.3/±13.2 ±5.9/±5.9 ±12.3/±14.0 ±4.6/±4.7
7 o′clock ±15.7/±14.0 ±5.9/±5.8 ±12.5/±12.4 ±5.7/±5.7 ±14.2/±13.7 ±4.7/±4.7
8 o′clock ±8.6/±8.4 ±4.6/±4.6 ±7.3/±9.7 ±4.3/±5.1 ±9.9/±9.2 ±4.4/±4.2
9 o′clock ±6.9/±7.4 ±4.1/±4.2 ±5.7/±7.0 ±3.6/±4.2 ±4.9/±8.6 ±3.0/±4.1
10 o′clock ±9.2/±8.7 ±4.8/±4.7 ±10.6/±8.6 ±5.3/±4.8 ±8.2/±10.0 ±4.0/±4.4
11 o′clock ±13.8/±12.5 ±5.7/±5.5 ±12.9/±14.0 ±5.8/±6.0 ±12.7/±14.4 ±4.7/±4.7
Table 3
 
Evaluation of cpRNFL Thickness Corrected With Littmann's and Kang's Methods
Table 3
 
Evaluation of cpRNFL Thickness Corrected With Littmann's and Kang's Methods
Sector Pearson r Bland–Altman Analysis
95% LoA Fixed Bias Proportional Bias
Sample Population Mean Difference (95% CI) Slope P Value
Global 0.940** ±6.0 ±4.4 −0.4 (−1.0 to 0.1) 0.005 0.87
Superior 0.882** ±14.3 ±8.4 −0.4 (−1.6 to 0.9) 0.016 0.72
Nasal 0.903** ±13.3 ±8.0 −1.2 (−2.4 to −0.1)* 0.045 0.26
Inferior 0.851** ±14.9 ±8.7 −0.2 (−1.5 to 1.2) −0.099 0.07
Temporal 0.880** ±16.8 ±9.3 0.1 (−1.4 to 1.5) 0.000 0.99
12 o′clock 0.895** ±18.8 ±9.9 −0.5 (−2.2 to 1.1) −0.010 0.79
1 o′clock 0.873** ±19.0 ±10.0 −0.7 (−2.4 to 1.0) 0.037 0.42
2 o′clock 0.885** ±16.1 ±9.1 −1.4 (−2.8 to 0.0) −0.016 0.72
3 o′clock 0.860** ±15.1 ±8.7 −1.8 (−3.1 to −0.5)* 0.071 0.08
4 o′clock 0.902** ±13.6 ±8.2 −1.6 (−2.8 to −0.4)* 0.055 0.17
5 o′clock 0.910** ±15.1 ±8.7 −1.5 (−2.8 to −0.1)* 0.000 0.99
6 o′clock 0.876** ±20.9 ±10.5 1.0 (−0.9 to 2.8) −0.134* <0.01
7 o′clock 0.825** ±27.5 ±11.8 0.4 (−2.0 to 2.8) −0.083 0.13
8 o′clock 0.851** ±23.7 ±11.2 0.4 (−1.7 to 2.5) 0.015 0.77
9 o′clock 0.902** ±12.6 ±7.8 0.4 (−0.7 to 1.6) 0.053 0.18
10 o′clock 0.913** ±18.6 ±9.9 0.0 (−1.6 to 1.6) −0.056 0.14
11 o′clock 0.731** ±27.9 ±11.9 0.0 (−2.5 to 2.4) 0.014 0.84
Table 4
 
Number of Eyes for Which the Difference in the Corrected cpRNFL Thickness by Littmann's and Kang's Methods Was Within the Measurement Error
Table 4
 
Number of Eyes for Which the Difference in the Corrected cpRNFL Thickness by Littmann's and Kang's Methods Was Within the Measurement Error
Sector Maximum Value of 95% LoA, μm No. of Eyes (%)
Global 5.2 121 (91.7)
Superior 9.6 119 (90.2)
Nasal 10.8 119 (90.2)
Inferior 9.8 115 (87.1)
Temporal 9.4 123 (93.2)
12 o′clock 12.1 113 (85.6)
1 o′clock 12.0 117 (88.6)
2 o′clock 13.8 118 (89.4)
3 o′clock 13.3 119 (90.2)
4 o′clock 13.1 124 (93.9)
5 o′clock 12.1 115 (87.1)
6 o′clock 13.1 110 (83.3)
7 o′clock 15.7 112 (84.8)
8 o′clock 9.9 121 (91.7)
9 o′clock 8.6 117 (88.6)
10 o′clock 10.6 119 (90.2)
11 o′clock 14.4 111 (84.1)
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×