Anita Garas

Reproducibility of Retinal Nerve Fiber Layer and Macular Thickness Measurement with the RTVue-100 Optical Coherence Tomograph

PURPOSE To evaluate the reproducibility of peripapillary retinal nerve fiber layer thickness (RNFLT) and ganglion cell complex (GCC) measurements made with the RTVue-100 Fourier-domain optical coherence tomography (OCT) device (Optovue, Inc., Fremont, CA) and to determine the influence of 4 factors: pupil dilation, subject age, experience in imaging examinations, and glaucoma severity. DESIGN Prospective study for evaluation of a diagnostic test. PARTICIPANTS Thirty-seven hospital-based participants (14 normal and ocular hypertensive subjects, 11 patients with moderate and 12 with severe glaucoma), all experienced in imaging examinations, and 40 consecutive screening trial participants lacking such experience. METHODS One eye of all participants was imaged 5 times. For the hospital-based participants, on the same day the measurement series was repeated after pupil dilation. For determination of intersession reproducibility, measurements were performed again 3 months later. MAIN OUTCOME MEASURES Intrasession and intersession coefficient of variation (CV), intraclass correlation coefficient (ICC), intratest variability, and test-retest variability. RESULTS For the average and quadrant RNFLT parameters and the GCC parameters, intrasession ICC varied between 93.9% and 99.0%, intrasession CV between 1.95% and 5.69%, and intratest variability varied between 3.11 and 9.13 microm. Most thickness values, all intrasession CV and intratest variability values, and the signal strength index were not changed significantly after pupil dilation. Intrasession CV of the 16 peripapillary RNFLT sectors varied between 4.90% and 11.66%. Most intrasession CV values increased significantly with increased disease severity, but intratest variability for average RNFLT and the quadrant RNFLT parameters, which are used for statistical comparison with the normative database, were unaffected by the severity of glaucoma. Patient age and experience in imaging examinations had no influence on intrasession CV. Intratest variability and intrasession CV represented 79.1% to 98.6% and 77.1% to 95.0% of test-retest variability and intervisit CV, respectively, for the average and quadrant RNFLT and the GCC parameters. CONCLUSIONS Intrasession and intersession reproducibility of RNFLT and GCC measurements with the RTVue-100 OCT are satisfactory for diagnostic purposes, both in normals and in patients with different glaucoma severity. Pupil dilation, age, and experience in imaging examinations did not influence reproducibility in a clinically significant manner. FINANCIAL DISCLOSURE(S) Proprietary or commercial disclosure may be found after the references.

Diagnostic accuracy of nerve fibre layer, macular thickness and optic disc measurements made with the RTVue-100 optical coherence tomograph to detect glaucoma.

PurposeTo evaluate the diagnostic accuracy of retinal nerve fibre layer thickness (RNFLT), ganglion cell complex (GCC), and optic disc measurements made with the RTVue-100 Fourier-domain optical coherence tomography (OCT) to detect glaucoma in a Caucasian referral population.MethodsOne randomly selected eye of 286 Caucasian patients (93 healthy, 36 ocular hypertensive, 46 preperimetric glaucoma, and 111 perimetric glaucoma eyes) was evaluated.ResultsUsing the software-provided classification, for the total population sensitivity did not exceed 73.6% for the optic nerve head parameters, and 62.7% for the other parameters. Specificity was high (94.6–100%) for most RNFLT and GCC parameters, but low (72.0–76.3%) for the optic disc parameters. Positive predictive value varied between 98.1 and 100% for the main RNFLT parameters, 92.6 and 100% for the 16 RNFLT sectors, 92.4 and 99.0% for the GCC parameters, but did not exceed 86.3% for any of the optic disc parameters. Positive likelihood ratio (PLR) was higher than 10 for average, inferior and superior RNFLT (25.5 to infinite), 12 of the 16 RNFLT sectors (12.6 to infinite), and three of the four GCC parameters (40.0 to 48.6). No optic disc parameter had a PLR higher than 3.0.ConclusionsRNFLT and GCC parameters of the RTVue-100 Fourier-domain OCT showed moderate sensitive but high specificity, positive predictive value and PLR for detection of glaucoma. The optic disc parameters had lower diagnostic accuracy than the RNFLT and GCC parameters.

Detection of early glaucomatous progression with different parameters of the RTVue optical coherence tomograph

Purpose:To investigate the ability of different parameters of the RTVue-100 Fourier-domain optical coherence tomograph (RTVue-OCT) to detect early glaucomatous progression. Methods:One eye of 17 healthy and 51 perimetric glaucoma patients was imaged prospectively at 6-month intervals for 1.5 to 3 years. Progression was determined by Octopus normal G2 visual field progression criteria. Results:Ten of the 51 glaucoma eyes progressed based on visual field criteria. Median visual field mean defect change was −0.300 dB/y for the controls, −0.120 dB/y for all glaucoma eyes (P=0.461 vs. controls), and 1.231 dB/y for the 10 functionally progressing glaucoma eyes (P<0.001 vs. controls). Ganglion cell complex (GCC), focal loss volume, and GCC global loss volume showed significantly faster rate of progression in the glaucoma group than in controls (P=0.004 and P=0.001, respectively). No optic nerve head, retinal nerve fiber layer thickness, and average GCC parameter separated the rate of progression between the groups. Conclusions:Early structural progression of glaucoma may be better detected with pattern-based GCC parameters of the RTVue-OCT than with the optic nerve head, retinal nerve fiber layer thickness or average GCC parameters of the same instrument.

Comparison of repeatability of retinal nerve fiber layer thickness measurement made using the rtvue fourier-domain optical coherence tomograph and the gdx scanning laser polarimeter with variable or enhanced corneal compensation

PurposeTo compare repeatability of measurements of peripapillary retinal nerve fiber layer thickness (RNFLT) made using the RTVue-100 Fourier-domain optical coherence tomograph against repeatability of those made using scanning laser polarimetry with variable corneal compensation or enhanced corneal compensation (GDx-VCC and GDx-ECC, respectively). MethodsOne eye of each of 37 participants (14 normal and ocular hypertensive subjects, 11 patients with moderate, and 12 with severe glaucoma; groups 1, 2, and 3, respectively) was imaged using the RTVue Optic Nerve Head Map scan, GDx-VCC, and GDx-ECC, each 5 times on the same day. The coefficient of variation (CV) were compared. The P values <1% were considered as significant. ResultsFor average RNFLT CV was significantly lower with RTVue (2.11%) than with GDx-ECC (3.22%, P=0.004), for all participants. For temporal quadrant RNFLT in all participants, and group 1, CV with RTVue (4.88% and 3.30%) was significantly lower than with GDx-ECC (7.40% and 5.88%; P=0.004), and tended to be lower than with GDx-VCC (6.81% and 5.80%; P=0.011 and 0.016, respectively). For all participants, CV for inferior quadrant RNFLT was significantly lower with RTVue (3.49%) than with GDx-VCC (5.20%, P=0.002). No other difference was seen for any other parameter in any subject group. ConclusionsRepeatability characterized by CV of RNFLT measurement with the RTVue Optic Nerve Head Map scan and GDx-VCC/GDx-ECC was similar, and sufficient for clinical purposes, both in normal subjects and in glaucoma patients with different disease severity. For average peripapillary RNFLT and temporal quadrant RNFLT, repeatability of RTVue was better than that of GDx-ECC, and tended to be better than that of GDx-VCC.

Structure-function relationship between the octopus perimeter cluster mean sensitivity and sector retinal nerve fiber layer thickness measured with the RTVue optical coherence tomography and scanning laser polarimetry

Purpose:To determine structure-function relationship between each of 16 Octopus perimeter G2 program clusters and the corresponding 16 peripapillary sector retinal nerve fiber layer thickness (RNFLT) values measured with the RTVue-100 Fourier-domain optical coherence tomography (RTVue OCT) and scanning laser polarimetry with variable corneal compensation (GDx-VCC) and enhanced corneal compensation (GDx-ECC) corneal compensation. Methods:One eye of 110 white patients (15 healthy, 20 ocular hypertensive, and 75 glaucoma eyes) were investigated. The Akaike information criterion and the F test were used to identify the best fitting model. Results:Parabolic relationship with logarithmic cluster mean sensitivity and linear sector RNFLT values provided the best fit. For RTVue OCT, significant (P<0.0001) coefficients of determination (R2) were found for all 16 RNFLT sectors. The R2 values were highest for the temporal, superotemporal, and inferotemporal RNFLT sectors (0.4483 to 0.5186). For GDx-VCC/ECC, significant (P<0.01) parabolic relationship was seen for all but the temporal and nasal RNFLT sectors. The overall highest R2 value (0.6943) was found for a superotemporal RNFLT sector with GDx-ECC. For some RNFLT sectors, the goodness of fit differed significantly between the imaging methods. Structure-function relationship was similar for the total population and the glaucoma subgroup, whereas no relationship (P>0.05) was found for the control eyes. Conclusions:Mean sensitivity of the Octopus visual field clusters showed significant parabolic relationship with the corresponding peripapillary RNFLT sectors. The relationship was more general with the RTVue OCT than GDx-VCC or GDx-ECC. The results show that visual field clusters of the Octopus G program can be applied for detailed structure-function research.

Novel Fractal Feature-Based Multiclass Glaucoma Detection and Progression Prediction

We investigate the use of fractal analysis (FA) as the basis of a system for multiclass prediction of the progression of glaucoma. FA is applied to pseudo 2-D images converted from 1-D retinal nerve fiber layer data obtained from the eyes of normal subjects, and from subjects with progressive and nonprogressive glaucoma. FA features are obtained using a box-counting method and a multifractional Brownian motion method that incorporates texture and multiresolution analyses. Both features are used for Gaussian kernel-based multiclass classification. Sensitivity, specificity, and area under receiver operating characteristic curve (AUROC) are computed for the FA features and for metrics obtained using wavelet-Fourier analysis (WFA) and fast-Fourier analysis (FFA). The AUROCs that predict progressors from nonprogressors based on classifiers trained using a dataset comprised of nonprogressors and ocular normal subjects are 0.70, 0.71, and 0.82 for WFA, FFA, and FA, respectively. The correct multiclass classification rates among progressors, nonprogressors, and ocular normal subjects are 0.82, 0.86, and 0.88 for WFA, FFA, and FA, respectively. Simultaneous multiclass classification among progressors, nonprogressors, and ocular normal subjects has not been previously described. The novel FA-based features achieve better performance with fewer features and less computational complexity than WFA and FFA.

Influence of age-related macular degeneration on macular thickness measurement made with fourier-domain optical coherence tomography

Purpose:To evaluate the influence of age-related macular degeneration (AMD) on macular thickness measurement made with Fourier-domain optical coherence tomography (RTVue-OCT) to detect glaucoma. Methods:One nonglaucomatous eye of 79 white persons was imaged. This comprised 25 healthy eyes, 19 eyes with early/intermediate AMD (geographic atrophy excluded), 16 eyes with subfoveal choroidal neovascularization (CNV), and 19 CNV eyes after intravitreal antiangiogenic treatment [CNV-antivascular endothelial growth factor (VEGF)]. Results:Compared with the age-matched controls, no difference in any nerve fiber layer and optic disc parameter was seen for any AMD group. No macular retinal segmentation error was detected in the control group. Localized inner retinal image segmentation errors topographically related to AMD were detected in 8 eyes with drusen (42.1%), all 16 CNV eyes (100%) and 17 eyes in the CNV–anti-VEGF group (89.5%; &khgr;2 test, P<0.001 for all comparisons). The average macular thickness parameters did not differ between the control and the AMD groups (analysis of variance, P>0.05). In contrast, all pattern-based ganglion cell complex (GCC) parameters were significantly higher (more abnormal) in the CNV and CNV–anti-VEGF group than in the control eyes (Mann-Whitney test, Bonferroni correction, P<0.001). For GCC focal loss volume, the only pattern-based parameter classified by the software, the frequency of “borderline” and “outside normal limits” classifications was significantly greater in each AMD group than in the control group (&khgr;2 test, Bonferroni correction, P⩽0.03). Conclusions:In nonglaucomatous eyes, AMD significantly influences the pattern-based inner macular thickness parameters of the RTVue optical coherence tomograph and the software-provided classification of GCC focal loss volume, for detection of glaucoma.

Comparison of diagnostic accuracy of the RTVue Fourier-domain OCT and the GDx-VCC/ECC polarimeter to detect glaucoma.

Purpose To compare sensitivity and specificity of retinal nerve fiber layer thickness (RNFLT) measurements made using RTVue-100 Fourier-domain optical coherence tomography (RTVue-OCT) and scanning laser polarimetry with variable (GDx-VCC) or enhanced compensation (GDx-ECC). Methods One eye of each of 177 consecutive patients was imaged. Healthy (n=50) and ocular hypertensive (n=28) eyes were defined as structurally undamaged, preperimetric (n=33) and perimetric (n=66) glaucoma eyes as diseased. Results For average RNFLT, sensitivity was higher (χ2 test, p=0.002) with RTVue-OCT (65.7%) than with GDx-VCC (49.5%). For superior and inferior RNFLT, sensitivity was similar with all methods. For the different nerve fiber bundle parameters, sensitivity of RTVue-OCT (64.6% to 84.8%) was consistently up to 35% higher (p<0.001) than that of GDx-VCC/ECC (28.3% to 72.7%). Specificity ranged from 84.6% to 98.7% with RTVue-OCT, 92.3% to 100% with GDx-VCC, and 94.9% to 100% with GDx-ECC, with no significant difference between the methods except for one nerve fiber bundle parameter, for which RTVue-OCT was less specific than either GDx method (p≤0.004). Diagnostic accuracy of the GDx-VCC/ECC nerve fiber indicator (NFI) and RTVue-OCT average RNFLT were similar. Of the detected glaucoma cases, 87.7% were identified both by GDx-VCC/ECC NFI and average RNFLT of RTVue-OCT. Conclusions In this clinical setting, all methods were similarly highly specific, but for localized RNFLT damage RTVue-OCT was statistically and clinically significantly more sensitive than GDx-VCC and GDx-ECC. Most detected glaucoma cases were identified with all 3 methods.

Automatic, operator-adjusted, and manual disc-definition for optic nerve head and retinal nerve fiber layer measurements with the RTVue-100 optical coherence tomograph.

PurposeTo evaluate the repeatability and between-methods differences of optic nerve head and retinal nerve fiber layer thickness (RNFLT) measurements as made with the automatic, operator-adjusted, and manual disc-definition functions of the RTVue-100 Fourier-domain optical coherence tomograph. MethodsOne eye of each of 126 persons (72 glaucomatous, 10 ocular hypertensive, 44 healthy control eyes) was analyzed with all 3 disc-definition functions. Thirty-three eyes were imaged 5 times for the determination of measurement repeatability. ResultsFor several parameters, repeatability with automatic disc definition was significantly poorer than the other options. In contrast, for all but 1 parameter repeatability did not differ between the manual and operator-adjusted methods. Refractive error, disease category or peripapillary atrophy had no influence on repeatability. However, the measured values differed significantly between the methods, for all but 3 parameters. Similar between-method differences were seen for eyes irrespective of glaucoma, refractive error, or parapapillary atrophy. High intraclass correlation (93.5% to 99.9%) was seen for all parameters measured with manual and operator-adjusted methods, but in the other comparisons only for RNFLT parameters. Agreement of the classifications was better for RNFLT parameters than disc parameters, and the &kgr;-values were highest (0.881 to 0.968) in comparing the manual and operator-adjusted methods. ConclusionsMeasurement repeatability with manual and operator-adjusted disc-definition functions is similar and is not influenced by refractive error, glaucoma, or parapapillary atrophy. The high intraclass correlation and the agreement of diagnostic classifications suggest that operator-adjusted disc-definition can be useful for glaucoma detection with the RTVue-100 Fourier-domain optical coherence tomograph.

Nerve fiber layer and macular thinning measured with different imaging methods during the course of acute optic neuritis.

Purpose TO compare retinal nerve fiber layer thickness (RNFLT) and inner macula thickness changes measured with Fourier-domain optical coherence tomography (FD-OCT) and scanning laser polarimetry during the course of acute optic neuritis (ON). Methods Nine eyes of 7 consecutive patients with multiple sclerosis (MS) were prospectively imaged from the onset of ON for 6 to 12 months. Nine healthy eyes were imaged for 12 to 19 months. Results Retinal nerve fiber layer thickness measured with FD-OCT initially increased in all eyes with diffuse optic disc edema. Inner macula thickness and Polarimetric RNFLT decreased already in the acute phase, in all eyes. All parameters stabilized at 2 to 5 months. The relative structural loss was different with the different Methods. Poor image quality with polarimetry occurred in 2 eyes in the acute phase of ON. In the control eyes all parameters were stable. Conclusions Change of RNFLT and macular thickness during the course of acute ON in MS strongly depends on the method used for the measurement. Inner macula thickness, measured with FD-OCT, was especially useful for the follow-up, since it was not influenced by initial disc edema and had consistently high image quality.