correlation between the ganglion cell-inner plexiform

8/21/2019 Correlation Between the Ganglion Cell-Inner Plexiform

http://slidepdf.com/reader/full/correlation-between-the-ganglion-cell-inner-plexiform 1/6

Glaucoma

Correlation Between the Ganglion Cell-Inner Plexiform Layer Thickness Measured With Cirrus HD-OCT andMacular Visual Field Sensitivity Measured With Microperimetry

Shino Sato, Kazuyuki Hirooka, Tetsuya Baba, Kaori Tenkumo, Eri Nitta, and Fumio Shiraga

Department of Ophthalmology, Kagawa University Faculty of Medicine, Miki, Kagawa, Japan

Correspondence: Kazuyuki Hirooka,Department of Ophthalmology, Ka-gawa University Faculty of Medicine,1750-1 Ikenobe, Miki, Kagawa 761-0793, Japan;[email protected].

Submitted: October 19, 2012 Accepted: March 30, 2013

Citation: Sato S, Hirooka K, Baba T,Tenkumo K, Nitta E, Shiraga F. Corre-lation between the ganglion cell-inner plexiform layer thickness measured with Cirrus HD-OCT and macular field visual sensitivity measured with mi-croperimetry. Invest Ophthalmol VisSci. 2013;54:3046–3051. DOI: 10.1167/iovs.12-11173

P URPOSE. To evaluate relationships between the macular visual field (VF) mean sensitivity andthe ganglion cell and inner plexiform layer (GCA) thicknesses.

METHODS. Seventy-one glaucoma patients and 29 healthy subjects were included in this cross-sectional study. At each visit, GCA thicknesses were measured by Cirrus HD-OCT and staticthreshold perimetry was performed using Macular Integrity Assessment (MAIA). Therelationship between the VF sensitivity and GCA thickness was examined globally, and inthe superior hemiretina, inferior hemiretina, and six VF sectors with both VF and optical

coherence tomography (OCT) in retinal view. Regression analysis was used to investigate therelationship between the GCA thickness and macular sensitivity.

R ESULTS. Macular VF sensitivity (dB) and GCA thickness relationships were statistically significant in each sector ( R ¼ 0.365–0.706, all P < 0.001). The highest correlation observed

was between the inferotemporal average mean sensitivity and the inferotemporal averageGCA thickness ( R ¼ 0.706) with both VF and OCT in retinal view. Strength of the structure–function relationship for each of the corresponding inferior sectors was higher than those for the corresponding superior sectors. The strength of the structure–function relationship of thetemporal sector was higher than that of the nasal sector.

CONCLUSIONS. GCA thickness measured by Cirrus HD-OCT showed statistically significantstructure–function associations with central VF. Inferotemporal central VF had the strongestassociation.

Keywords: optical coherence tomography, visual field, microperimetry

A dvances of optical coherence tomography (OCT) have

enabled assessment of retinal ganglion cell (RGC) axons by

measuring the thickness of the peripapillary retinal nerve fiber layer (RNFL) and the macular area. More recent advances in

segmentation algorithms have it made possible to use OCT to

visualize and measure individual retinal layers in the macular region.1–3 RTVue-100 OCT (Optovue, Inc., Fremont, CA)

incorporates a ganglion cell complex (GCC) scan mode thatmeasures the inner macular retinal layer thickness from the

internal limiting membrane to the inner plexiform layer (IPL), which is composed of ganglion cell axons, cell bodies, and

dendrites. Previous studies have shown that the macular GCC

thickness measurements derived from GCC scan data are

significantly lower in glaucomatous eyes with visual field (VF)defects than in healthy eyes, and have a good glaucoma

discriminating power that is comparable to that of the RNFL.4–7

The thickness of the RGC or RGC þ IPL (GCA: ganglion cell

analysis) in the macula has also been measured by OCT.2,8

Mwanza et al.9,10 recently showed that the Cirrus HD-OCT GCA

algorithm (Carl Zeiss Meditec, Dublin, CA) can successfully detect and measure the inner macular layers (the GCA; an area

that contains the ganglion cell layer and the IPL) with excellentintervisit reproducibility.

Microperimetry, which is known as fundus controlled

perimetry or fundus perimetry, assesses retinal sensitivity during the direct examination of the ocular fundus. Micro-

perimetry data are independent of eye movements and exactly related to the stimulated area. In addition, VF sensitivity can

also be measured by microperimetry with improved spatiallocalization.

A number of studies have used spectral-domain OCT tofocus on the relationship between structural and functional

damage as a way to improve our ability to detect the presenceand progression of glaucomatous damage. Results of thesestudies have demonstrated there are high correlations between

the global VF sensitivity and the peripapillary RNFL/GCC

thickness.5,11,12

The macular VF, including the central vision, is very important if a glaucoma patient is to enjoy normal daily life. Therefore, preservation of the macular VF is the key

concern in glaucoma management. Wang et al.2 found that localthickness of the RGC þ IPL (GCA) could be obtained from

frequency-domain OCT scans, with these measurementsshowing qualitative agreement with local VF sensitivity. Raza

et al.8 recently showed a strong relationship between losses instandard automated perimetry (SAP) sensitivity and decreasesin local RGC þ IPL thicknesses (GCA).

Copyright 2013 The Association for Research in Vision and Ophthalmology, Inc.

www.iovs.org j ISSN: 1552-5783 3046



The purpose of the current study was to evaluate thecorrelation between the GCA thickness and the local VFsensitivity obtained when using microperimetry, Macular Integrity Assessment (MAIA; CenterVue, Padova, Italy), in each sector of the macula.

M ATERIALS AND METHODS

All patients were examined at Kagawa University Hospital fromNovember 2011 through May 2012. At each visit, the GCA algorithm was used to detect the macular inner structurethickness, while MAIA was used to determine the staticthreshold perimetry. All eligible subjects received a detailedexplanation of the study and signed an informed consent formin accordance with the principles embodied in the Declarationof Helsinki. The study protocol was approved by theinstitutional review board of the Kagawa University Faculty of Medicine. Healthy control subjects were either subjectsattending the outpatients clinics, spouses and friends of therecruited patients, or volunteers from the hospital staff.

All subjects underwent a complete ophthalmic examinationthat included visual acuity testing with refraction, IOP,gonioscopy examinations, and dilated fundus examination

with stereoscopic biomicroscopy of the optic nerve headusing slit lamp and indirect ophthalmoscopy. To be included inthe study, all subjects had to have a best-corrected visual acuity of 20/40 or better, a spherical error within a range betweenþ4.0 and 6.0 diopters (D), a cylinder within 6 2.0 D, andopen angles (grade 3 and 4 according to the Shaffer gradingsystem). Exclusion criteria included a history of any kind of retinal pathology or neurologic disease, retinal laser procedure,or either retinal or intraocular surgery. One eye in each subject

was randomly chosen for inclusion in the study. To be enrolledas a control in the study, subjects had to have an IOP less thanor equal to 21 mm Hg, no history of retinal pathology, and anormal visual field. Glaucomatous eyes were defined as eyesexhibiting structural glaucomatous changes (vertical cup-discasymmetry between fellow eyes of ‡ 0.2, a cup-to-disc ratio of

‡ 0.6, and neuroretinal rim narrowing, notches, localizedpallor, or RNFL defects with glaucomatous VF loss in thecorresponding hemifield). A glaucomatous VF was defined as aglaucoma hemifield test (GHT) outside normal limits on at leasttwo consecutive baseline tests and the presence of at leastthree contiguous test points within the same hemifield on thepattern deviation plot at P less than 1%, with at least one at P less than 0.5% excluding points on the edge of the field or those directly above and below the blind spot.

Cirrus HD-OCT Imaging

All eyes were scanned by the Cirrus HD-OCT system. Only good-quality scans were used for the analyses. To be included,scans had to have a signal strength less than or equal to 6, be

without RNFL discontinuity or misalignment, involuntary saccade or blinking artifacts, and show an absence of algorithmsegmentation failure during a careful visual inspection. The GCanalysis algorithm was used to automatically measure themacular GCA thickness. Software version 6.0 of the GCA algorithm was used to process the data in this study, as it wasable to detect and measure the thickness of the macular ganglion cell-inner plexiform layer within a 14.13 mm2

elliptical annulus area centered on the fovea. The protocolcarries out 200 horizontal B-scans, which are comprised of 200

A-scans per B-scan that are performed 1024 times within acube measuring 6 3 6 3 2 mm. These scans were designed toallow analysis of the retinal topography.8,9 The GCA algorithmis able to process data from either of the three-dimensional

(3D) volume scans performed by the Cirrus HD-OCT. Both of these scan patterns cover the same physical field of view,namely an area that is 6 3 6 3 2 mm. However, thedimensionality of the image data is either 512 3 128 3 1024or 200 3 200 3 1024. The input image data are initially segmented using the existing Cirrus inner limiting membrane(ILM) and RPE segmentation algorithms in order to create aregion of interest that lies within the intraretinal layers.9,10 Thealgorithm identifies the RNFL layer, the GCA layer (an area thatranges from the outer boundary of the RNFL to the outer boundary of the IPL, and that includes both the RGC layer andthe IPL), and the outer retinal layer. The segmentationprocedure operates in three dimensions and uses a graph-based algorithm to identify each layer. A detailed description of how the algorithm operates has been previously presented indetail.9,10 There are nine thicknesses measured within theannular area that is centered on the fovea. These include theaverage, superior average, inferior average, and six sectoral(superotemporal, superior, superonasal, inferotemporal, inferi-or, inferonasal) values of the respective layers.

Microperimetry Examinations

In all subjects, MAIA was used to measure the retinal

sensitivity. MAIA was performed in a dim room without any dilation of the pupil. The following parameters were used inthe current study: a 68-stimuli grid covering the central 108 of the retina, a 4 to 2 threshold strategy, a fixation target thatconsisted of a red circle with a 18 diameter, stimulus sizeGoldmann III, background luminance set at 4 apostilb (asb),maximum luminance of 1000 asb, and a stimulus dynamicrange of 36 dB. Use of the MAIA device made it possible todetermine which of the fixation stabilities were stable. Only stable fixation tests were included in the analysis.

Mapping Structure to Function

For analysis of functional measurements, VF sensitivity wasobtained for each subject. According to the criteria described

earlier, 8 of 108 eyes that qualified for initial inclusion wereexcluded. Seven of these eyes had a signal strength less than 6,

while one eye showed unstable fixation. Therefore, a total of 100 eyes were included in the final analysis. When comparingthe GCA thickness to the local loss in VF sensitivity, it isimportant that the displacement of the RGCs in the macula betaken into consideration.13 The average location of the RGCsassociated with each MAIA test point was approximated usingequations that were derived from the histologic analysis-based

work of Drasdo et al.14 Figure 1 shows the location of the visual field MAIA test points, while Figure 2A shows thelocation after adjusting for the RGC displacement. Structure-function relationships were determined from each of the sixsectors with both VF and OCT in retinal view (Fig. 2B). Sincefundus perimetry data are exactly related to the stimulated

area, functional measurements of MAIA were correlated with the structural measurements of the OCT in the same hemifield.

Statistical Analysis

To determine the association between the local VF meansensitivity and the relative GCA thickness, we based this study on previous work that related the SAP sensitivity to peripap-illary RNFL thickness15–17 or RGCþ IPL thickness (GCA) in themacula.8 Briefly, the assumptions for this model were asfollows: the measured GCA thickness R consists of twocomponents, the thickness S , which is due to portions of theIPL and RGC layer (GCA) that is affected by glaucoma, and theresidual B, which includes portions of the IPL and RGC layer

Relationship Between Macular VF and GCA IOVS j April 2013 j Vol. 54 j No. 4 j 3047



(GCA) not affected by glaucoma. These areas may include glialcells, blood vessels, bipolar cell axons, and amacrine cellprojections. Thus, overall R ¼ S þ B. As the MAIA fieldsensitivity changes, the value of S will decrease, while B willremain constant. As the VF sensitivity decreases, the signalportion S of the local GCA thickness R will decrease linearly

when the VF sensitivity is expressed in linear units. Thus, R ¼( S 0 B ) T þ B for T less than or equal to 1.0, where S 0 is themedian of the control GCA thickness at a particular eccentric-ity, T is the relative sensitivity (defined as 100.1D ), and D is the

VF sensitivity minus the mean value of healthy subjects. T equals 1.0 when there is no loss (0 dB difference from normal)

and approaches 0 when there are large losses in the sensitivity.The variable B was calculated for each zone as the median of the GCA data when the local VF sensitivities were less than 20dB.

Correlations of the GCA thickness with the corresponding VF mean sensitivity were examined by using Spearman rank order correlations. Differences between the control andglaucoma groups were assessed by an independent Student’st -test and the v

2 test for categorical parameters. All statistical values are presented as the mean 6 SD, with P values less than

0.05 considered to be statistically significant. Statisticalanalyses were performed using SPSS version 19.0 (IBM, New

FIGURE 1. Detection of macular sensitivity. Right eye fundus image of a 50-year-old patient with primary open-angle glaucoma (POAG). Figureshows microperimetry results with differential light threshold values color-coded from 0 to 36 dB.

FIGURE 2. Representative example showing the six sectors of the ganglion cell analyzer thickness and the adjustment for retinal ganglion celldisplacement. ( A ) The 10 to 2 VF displaced points corresponding to the RGC locations based on a model derived from histologic analysis.5 ( B ) GCA thickness areas were divided into six sectors, with the corresponding VF regions then obtained using MAIA.




York, NY). Comparisons of the strength structure–functionassociation were evaluated by tests of equality of dependentcorrelation coefficients.

R ESULTS

A total of 71 glaucoma patients (32 POAG, 37 normal-tension,

and 2 exfoliative) and 29 healthy subjects were enrolled in thestudy. The demographic characteristics are presented in Table1. Since the healthy control participants were selected basedon age- and refractive error-matching, there was no significantdifference noted in the mean age between the healthy subjectsand glaucoma patients. Based on the standard VF severity grading scale,18 the grade of the disease in the glaucomatouseyes of the 71 patients ranged from early to moderate, with 29(41%) classified as early, 24 (34%) classified as moderate, and18 (25%) classified as severe.

Table 2 lists the GCA thickness, while Table 3 presents themacular sensitivity. GCA thickness and macular sensitivity weresignificantly different between the glaucoma and healthy subject groups.

Figure 3 shows the structure–function relationship be-

tween the GCA thickness and the corresponding VF meansensitivity with both OCT and VF in retinal view. To check thegoodness of fit, we calculated the number of points fallingoutside the 95% confidence boundary. For the global GCA thickness–VF mean sensitivity, seven eyes fell outside of the95% confidence boundary. Although the difference in thestructure–function relationship was not significant among sixsectors, the highest Spearman correlation coefficient was0.706 in the inferotemporal sector. Strength of the structure–function relationship of each of the corresponding inferior sectors ( R ¼ 0.546–0.706) were higher than those of the

corresponding superior sectors ( R ¼ 0.365–0.601). Inaddition, the strength of the structure–function relationshipof the temporal sector was higher than that of the nasal sector (inferonasal [ R ¼ 0.546] versus inferotemporal [ R ¼ 0.706] [ P

¼ 0.02], superonasal [ R ¼ 0.365] versus superotemporal [ R ¼0.601] [ P ¼ 0.002]).

DISCUSSION

The pathology of glaucoma is characterized by the death of RGCs and their axons. Although there is substantial individual

variability, the average retina contains 1.07 million RGCs, with approximately 50% of the RGCs located within 4.5 mm of thefovea.19,20 It has been demonstrated in humans that the RGCloss is evident around the fovea during the early stages of thedisease.21 Therefore, we selected the central macular VF as thelocal functional target in the current study and then assessed itsstructure–function relationship by using the GCA thickness.

Although SAP is able to show the functioning of individualretinal locations in the macula, reliable test results can be

difficult to obtain because of unstable fixation in some cases. When using an auto trackin g syst em, MAIA is able toautomatically record the fixation behavior during the test

while the autotracking system makes it possible to adjust thestimulus points to predefined retinal positions and performreliable field testing, even in eyes with unstable fixation.Because of these advantages, this study used MAIA to measurethe macular VF sensitivity.

The current study examined structure–function relation-ships in smaller regions, which to the best of our knowledge,has not been previously investigated. The determination of

which region has a stronger structure–function association hasclinical implications, as use of the stronger region couldprovide better detection and follow up in glaucoma patients

who may present with an early stage of macular VF defects.Thus, our current results indicate the importance of studyingthe relationship patterns in smaller regions. The strength of thestructure–function relationship is related to the individualanatomy and its variation in the subject, the stage of glaucomapresent in the study sample, the VF scale, and the regressionmodel used. During the previous examinations of thecorrelations between the VF sensitivity and the early glau-comatous stages, healthy individuals, and the glaucoma suspecteyes, the correlations were shown to be weaker than thoseobserved in moderate-to-severe glaucoma.22,23 The reason for this is because the range of VF is narrow in healthy andglaucoma suspect eyes. Thus, the strength of a structure–function relationship may be dependent upon both the actualretinal and VF areas in which the association is assessed and

T ABLE 1. Clinical Characteristics of the Study Population

Glaucoma Normal P

Age, y 62.5 6 11.0 64.3 6 12.2 0.71

Sex (M/F) 40/31 17/12 0.83

Diagnosis

POAG 32

NTG 37

EG 2

Refraction (D) 1.6 6 2.3 0.7 6 1.9 0.10

NTG, normal-tension glaucoma; EG, exfoliation glaucoma; M, male;F, female; D, diopter.

T ABLE 2. GCA Thickness for the Average, Superior Average, Inferior Average, and in Each of the Six Sectors

GCA Thickness, lm

P Glaucoma Normal

Average 65.1 6 7.7 79.8 6 5.2 <0.01

Superior hemifield 67.1 6 9.2 80.8 6 5.5 <0.01

Inferior hemifield 63.1 6 8.0 78.8 6 5.5 <0.01

Superotemporal 63.9 6 9.8 79.6 6 5.1 <0.01

Superior 67.5 6 9.6 81.0 6 6.5 <0.01

Superonasal 70.9 6 10.0 81.7 6 6.3 <0.01

Inferotemporal 59.6 6 9.0 80.4 6 6.1 <0.01

Inferior 62.0 6 8.4 76.6 6 5.9 <0.01

Inferonasal 67.7 6 9.2 79.4 6 5.7 <0.01

T ABLE 3. Macular Sensitivity for the Average, Superior Average,Inferior Average, and in Each of the Six Sectors

Mean Sensitivity, dB

P Glaucoma Normal

Average 21.4 6 6.3 26.9 6 1.5 <0.01

Superior hemifield 23.4 6 6.0 26.9 6 1.6 <0.01

Inferior hemifield 19.4 6 8.4 26.9 6 1.5 <0.01

Superotemporal 21.4 6 8.3 26.9 6 1.7 <0.01

Superior 23.6 6 6.8 26.8 6 2.0 0.02

Superonasal 25.1 6 5.6 27.1 6 1.7 0.07

Inferotemporal 17.5 6 10.3 27.0 6 1.7 <0.01

Inferior 16.8 6 11.1 26.4 6 1.6 <0.01

Inferonasal 24.2 6 6.5 27.1 6 1.8 0.02




the specific imaging device that is used in the study. Inaddition, histologic studies in human24,25 and monkey eyes26

have shown that in the central retina, there are more ganglioncells in the nasal and superior sectors than in the temporal andinferior sectors, respectively. These differences in the distribu-tion of the ganglion cells might affect the strength of thestructure–function relationship.

Raza et al.8 recently reported finding that the local RGC þIPL thickness (GCA) correlated well with the local sensitivity loss obtained with the 10 to 2 SAP setting in the central 7.28 of the 14 patients with glaucoma and in the 19 healthy subjects.The differences between the current and previous study wereas follows: our study showed that the strength of thecorrelation between the VF mean sensitivity and the GCA thickness varied within the macular region. In addition, our study measured the VF sensitivity with microperimetry in anattempt to improve the spatial localization. We also examined atotal of 100 eyes, as it was hoped this larger sample size wouldresult in a greater statistical power.

In the past, relationships between RNFL losses and VFdefects have been studied using different theoretical curves tofit the data.15–17,27,28 Results have shown that a complete loss

in sensitivity does not result in a RNFL thickness of zero, butinstead, is actually associated with a finite RNFL thickness. Inthe early stage of glaucoma, the decline in RNFL thickness israpid, and there is a lag in the visual sensitivity loss. However,as the glaucoma becomes severe, RNFL thicknesses reach abase level beyond which only the visual sensitivity declines.For these reasons, we decided to evaluate the structure–function relationship using the model proposed by Hood etal.15,16 Our results showed that their linear model fit thestructure–function data quite well.

With regard to potential limitations, we were not able toobserve any overall age effect, the number of subjects in thisstudy was relatively small; however, a further study with alarger number of subjects should be able to address this issue.

We also did not notice any obvious difference in the structure-function agreement for patients with POAG, normal-tensionglaucoma, or exfoliation glaucoma. Additional studies thatmore closely examine the different types of glaucoma will needto be undertaken.

In conclusion, although the GCA thickness measured by Cirrus HD-OCT was significantly correlated with the macular retinal sensitivity assessed by MAIA, the strength of the

FIGURE 3. Scatters plots showing the association between the Cirrus HD-OCT thickness parameters and the corresponding retinal sensitivity. ( A ) Average thickness of the GCA versus the macular mean sensitivity. ( B ) Superior average thickness of the GCA versus the superior macular mean

sensitivity. ( C ) Inferior average thickness of the GCA versus the inferior macular mean sensitivity. ( D ) Superotemporal thickness of the GCA versusthe superotemporal mean sensitivity. ( E ) Superior thickness of the GCA versus the superior mean sensitivity. ( F ) Superonasal thickness of the GCA versus the superonasal mean sensitivity. ( G ) Inferotemporal thickness of the GCA versus the inferotemporal mean sensitivity. ( H ) Inferior thicknessof the GCA versus the inferior mean sensitivity. ( I ) Inferonasal thickness of the GCA versus the inferonasal mean sensitivity. Spearman correlationcoefficients, * P < 0.001.




correlation varied from region to region. Combining GCA thickness with macular VF sensitivity may provide a better understanding of the amount of glaucomatous damage thatoccurs in the macula region.

Acknowledgments

Disclosure: S. Sato, None; K. Hirooka , None; T. Baba , None; K. Tenkumo, None; E. Nitta , None; F. Shiraga , None

References

1. Ishikawa H, Stein DM, Wollstein G, et al. Macular segmentation with optical coherence tomography. Invest Ophthalmol VisSci . 2005;46:2012–2017.

2. Wang M, Hood DC, Cho JS, et al. Measurement of local retinalganglion cell layer thickness in patients with glaucoma usingfrequency-domain optical coherence tomography. Arch Oph- thalmol . 2009;127:875–881.

3. Mishra A, Wong A, Bizheva K, Clausi DA. Intra-retinal layer segmentation in optical coherence tomography images. Opt Express. 2009;17:23719–23728.

4. Garas A, Vargha P, Hollo G. Diagnostic accuracy of nerve fiber layer, macular thickness and optic disc measurements made

with the RTVue-100 optical coherence tomograph to detectglaucoma. Eye (Lond). 2011;25:57–65.

5. Kim NR, Lee ES, Seong GJ, et al. Structure-function relation-ship and diagnostic value of macular ganglion cell complexmeasurement using Fourier-domain OCT in glaucoma. Invest Ophthalmol Vis Sci . 2010;51:4646–4651.

6. Schulze A, Lamparter J, Pfeiffer N, et al. Diagnostic ability of retinal ganglion cell complex, retinal nerve fiber layer, andoptic nerve head measurements by Fourier-domain opticalcoherence tomography. Graefes Arch Clin Exp Ophthalmol .2011;249:1039–1045.

7. Tan O, Chopra V, Lu AT, et al. Detection of macular ganglioncell loss in glaucoma by Fourier-domain optical coherencetomography. Ophthalmology. 2009;116:2305–2314.

8. Raza AS, Cho J, de Moraes CG, et al. Retinal ganglion cell layer

thickness and local visual field sensitivity in glaucoma. ArchOphthalmol . 2011;129:1529–1536.

9. Mwanza JC, Durbin MK, Budenz DL, et al. Profile andpredictors of normal ganglion cell-inner plexiform layer thickness measured with frequency-domain optical coherencetomography. Invest Ophthalmol Vis Sci . 2011;52:7872–7879.

10. Mwanza JC, Oakley JD, Budenz DL, et al. Macular ganglion cell-inner plexiform layer: automated detection and thicknessreproducibility with spectral domain-optical coherence to-mography in glaucoma. Invest Ophthalmol Vis Sci . 2011;52:8323–8329.

11. Cho JW, Sung KR, Lee S, et al. Relationship between visualfield sensitivity and macular ganglion cell complex thicknessas measured by spectral-domain optical coherence tomogra-phy. Invest Ophthalmol Vis Sci . 2010;51:6401–6407.

12. Takagishi M, Hirooka K, Baba T, Mizote M, Shiraga F.Comparison of retinal nerve fiber layer thickness measure-

ments using time domain and spectral domain opticalcoherence tomography, and visual field sensitivity. J Glauco- ma. 2011;20:383–387.

13. Hood DC, Raza AS, de Moraes CG, et al. Initial arcuate defects within the central 10 degrees in glaucoma. Invest Ophthalmol Vis Sci . 2011;52:940–946.

14. Drasdo N, Millican CL, Katholi CR, Curcio CA. The length of Henle fibers in the human retina and a model of ganglionreceptive field density in the visual field. Vision Res. 2007;47:2901–2911.

15. Hood DC, Anderson SC, Wall M, Kardon RH. Structure versusfunction in glaucoma: an application of a linear model. Invest Ophthalmol Vis Sci . 2007;48:3662–3668.

16. Hood DC, Kardon RH. A framework for comparing structuraland functional measures of glaucomatous damage. Prog Retina Eye Res. 2007;26:688–710.

17. Horn FK, Mardin CY, Laemmer R, et al. Correlation betweenlocal glaucomatous visual field defects and loss of nerve fiber layer thickness measured with polarimetry and spectraldomain OCT. Invest Ophthalmol Vis Sci . 2009;50:1971–1977.

18. Hodapp E, Parrish RK II, Anderson DR. Clinical Decision inGlaucoma. St. Louis, MO: Mosby; 1993:52–61.

19. Curcio CA, Allen KA. Topography of ganglion cells in humanretina. J Comp Neurol . 1990;300:5–25.

20. Wassle H, Gr unert U, Rohrenbeck J, Boycott BB. Retinalganglion cell density and cortical magnification factor in theprimate. Vision Res. 1990;30:1897–1911.

21. Quigley HA, Dunkelberger GR, Green WR. Retinal ganglioncell atrophy correlated with automated perimetry in humaneyes with glaucoma. Am J Ophthalmol . 1989;107:453–464.

22. Miglior S, Riva I, Guareschi M, et al. Retinal sensitivity andretinal nerve fiber layer thickness measured by opticalcoherence tomography in glaucoma. Am J Ophthalmol .2007;144:733–740.

23. Ajtony C, Balla Z, Somoskeoy S, Kovacs B. Relationshipbetween visual field sensitivity and retinal nerve fiber thickness as measured by optical coherence tomography. Invest Ophthalmol Vis Sci . 2007;48:258–263.

24. Curcio CA, Allen KA. Topography of ganglion cells in humanretina. J Comp Neurol . 1990;300:5–25.

25. Curcio CA, Messinger JD, Sloan KR, et al. Human chorioretinallayer thickness measured using macula-wide high resolutionhistological sections. Invest Ophthalmol Vis Sci . 2011;52:3943–3954.

26. Perry VH, Cowey A. The ganglion cell and cone distributionsin the monkey’s retina: implications for central magnificationfactors. Vision Res. 1985;25:1795–1810.

27. Garway-Heath DF, Caprioli J, Fitzke FW, Hitchings RA. Scalingthe hill of vision: the physiological relationship between lightsensitivity and ganglion cell numbers. Invest Ophthalmol VisSci . 2000;41:1774–1782.

28. Garway-Heath DF, Holder GE, Fitzke FW, Hitchings RA.Relationship between electrophysiological, psychophysical,and anatomical measurements in glaucoma. Invest Ophthal- mol Vis Sci . 2002;43:2213–2220.


correlation between the ganglion cell-inner plexiform

Documents