Effect of spectrum bias on the diagnostic accuracy of spectral-domain optical coherence tomography in glaucoma

Optical Coherence Tomography and Optical Coherence Tomography Angiography: Essential Tools for Detecting Glaucoma and Disease Progression

Early diagnosis and detection of disease progression are critical to successful therapeutic intervention in glaucoma, the leading cause of irreversible blindness worldwide. Optical coherence tomography (OCT) is a non-invasive imaging technique that allows objective quantification in vivo of key glaucomatous structural changes in the retina and the optic nerve head (ONH). Advances in OCT technology have increased the scan speed and enhanced image quality, contributing to early glaucoma diagnosis and monitoring, as well as the visualization of critically important structures deep within the ONH, such as the lamina cribrosa. OCT angiography (OCTA) is a dye-free technique for noninvasively assessing ocular microvasculature, including capillaries within each plexus serving the macula, peripapillary retina and ONH regions, as well as the deeper vessels of the choroid. This layer-specific assessment of the microvasculature has provided evidence that retinal and choroidal vascular impairments can occur during early stages of glaucoma, suggesting that OCTA-derived measurements could be used as biomarkers for enhancing detection of glaucoma and its progression, as well as to reveal novel insights about pathophysiology. Moreover, these innovations have demonstrated that damage to the macula, a critical region for the vision-related quality of life, can be observed in the early stages of glaucomatous eyes, leading to a paradigm shift in glaucoma monitoring. Other advances in software and hardware, such as artificial intelligence-based algorithms, adaptive optics, and visible-light OCT, may further benefit clinical management of glaucoma in the future. This article reviews the utility of OCT and OCTA for glaucoma diagnosis and disease progression detection, emphasizes the importance of detecting macula damage in glaucoma, and highlights the future perspective of OCT and OCTA. We conclude that the OCT and OCTA are essential glaucoma detection and monitoring tools, leading to clinical and economic benefits for patients and society.

Role Of Adaptive Optics In Early Diagnosis Of Glaucoma From A Clinician's Perspective

BACKGROUND

Glaucoma is one of the leading causes of irreversible blindness across the world. Early detection is important to minimize the loss of visual function. The diagnostic tools, optical coherence tomography (OCT) and standard automated perimetry (SAP) form the keystones of the diagnosis and monitoring of the condition. However, the ability of these tools to diagnose early forms of glaucoma is limited. Adaptive optics (AO) is a technology that could help to overcome this limitation. AO technology can detect slightest changes occurring at the cellular level by compensating for ocular aberrations.

METHODS

We searched PubMed for publications between 2002 and 2019 on adaptive optics in Ophthalmology. The key words were adaptive optics, lamina cribrosa, retinal nerve fiber layer defects, scanning laser ophthalmoscope and OCT.

RESULTS

Out of 38 publications, 17 original articles or case series with relevance to glaucoma, and written in English were selected and reviewed.

CONCLUSIONS

The AO technology, combined with various platforms such as fundus photography, scanning laser ophthalmoscopy and OCT, has been used in glaucoma patients to study the lamina cribrosa, retinal nerve fiber layer (RNFL), retinal photoreceptors as well as ocular circulation in minute detail. Imaging the subtle changes in morphology and reflectivity of RNFL at the preclinical stage may lead to early detection of glaucoma. Longitudinal monitoring of RNFL alterations in glaucoma patients is possible. At present, the technology is expensive with limited availability, and has several limitations.

Adaptive optics scanning laser ophthalmoscopy may support early diagnosis of glaucoma

Purpose:

To compare image characteristics of retinal nerve fiber layer (RNFL) between glaucoma patients and healthy controls using adaptive optics scanning laser ophthalmoscopy (AOSLO).

Methods:

This was a cross-sectional pilot study with two groups: a glaucoma group with patients with moderate or severe glaucoma as per the Hodapp–Parrish–Anderson classification system and a control group with healthy individuals. The optic nerve damage in moderate glaucoma was predominantly located in only one hemisphere; the other hemisphere was un- or minimally affected on optical coherence tomography and automated perimetry and is referred to as early glaucoma. The structure of RNFL bundles and gain (%) in RNFL images with mean pixel values between 15 and 35 were analyzed. Imaging was performed one degree away from the optic disc margin at two and four cardinal clock positions in the glaucoma and control groups, respectively. The field of view was 1.3° at 2.3 m resolution. We studied one eye per participant.

Results:

There were 11 glaucoma patients and 7 healthy controls. Imaging was successful at 88% of the locations in controls and early glaucoma; the reflectivity differed significantly (0.51 and 0.56, respectively, P < 0.001) but not the structure of RNFL bundles (Cohen’s Kappa 0.11) between them. In patients with moderate and severe glaucoma, imaging was successful only at 46% of the locations; RNFL bundles were not discernible, and RNFL reflectivity did not differ from those with early glaucoma (P < 0.11).

Conclusion:

The recorded gain (%) of RNFL images obtained using AOSLO could be an objective indicator of early glaucoma.

Performances of Machine Learning in Detecting Glaucoma Using Fundus and Retinal Optical Coherence Tomography Images: A Meta-Analysis

PURPOSE

To evaluate the performance of machine learning (ML) in detecting glaucoma using fundus and retinal optical coherence tomography (OCT) images.

DESIGN

Meta-analysis.

METHODS

PubMed and EMBASE were searched on August 11, 2021. A bivariate random-effects model was used to pool ML's diagnostic sensitivity, specificity, and area under the curve (AUC). Subgroup analyses were performed based on ML classifier categories and dataset types.

RESULTS

One hundred and five studies (3.3%) were retrieved. Seventy-three (69.5%), 30 (28.6%), and 2 (1.9%) studies tested ML using fundus, OCT, and both image types, respectively. Total testing data numbers were 197,174 for fundus and 16,039 for OCT. Overall, ML showed excellent performances for both fundus (pooled sensitivity = 0.92 [95% CI, 0.91-0.93]; specificity = 0.93 [95% CI, 0.91-0.94]; and AUC = 0.97 [95% CI, 0.95-0.98]) and OCT (pooled sensitivity = 0.90 [95% CI, 0.86-0.92]; specificity = 0.91 [95% CI, 0.89-0.92]; and AUC = 0.96 [95% CI, 0.93-0.97]). ML performed similarly using all data and external data for fundus and the external test result of OCT was less robust (AUC = 0.87). When comparing different classifier categories, although support vector machine showed the highest performance (pooled sensitivity, specificity, and AUC ranges, 0.92-0.96, 0.95-0.97, and 0.96-0.99, respectively), results by neural network and others were still good (pooled sensitivity, specificity, and AUC ranges, 0.88-0.93, 0.90-0.93, 0.95-0.97, respectively). When analyzed based on dataset types, ML demonstrated consistent performances on clinical datasets (fundus AUC = 0.98 [95% CI, 0.97-0.99] and OCT AUC = 0.95 [95% 0.93-0.97]).

CONCLUSIONS

Performance of ML in detecting glaucoma compares favorably to that of experts and is promising for clinical application. Future prospective studies are needed to better evaluate its real-world utility.

Measurements of OCT Angiography Complement OCT for Diagnosing Early Primary Open-Angle Glaucoma

PURPOSE

To compare measurements of global and regional circumpapillary capillary density (cpCD) with retinal nerve fiber layer (RNFL) thickness and characterize their relationship with visual function in early primary open-angle glaucoma (POAG).

DESIGN

Cross-sectional study.

PARTICIPANTS

Eighty healthy eyes, 64 preperimetric eyes, and 184 mild POAG eyes from the Diagnostic Innovations in Glaucoma Study.

METHODS

Global and regional RNFL thickness and cpCD measurements were obtained using OCT and OCT angiography (OCTA). For direct comparison at the individual and diagnostic group level, RNFL thickness and capillary density values were converted to a normalized relative loss scale.

MAIN OUTCOME MEASURES

Retinal nerve fiber layer thickness and cpCD normalized loss at the individual level and diagnostic group. Global and regional areas under the receiver operating characteristic curve (AUROC) for RNFL thickness and cpCD to detect preperimetric glaucoma and glaucoma, R² for the strength of associations between RNFL thickness function and capillary density function in diagnostic groups.

RESULTS

Both global and regional RNFL thickness and cpCD decreased progressively with increasing glaucoma severity (P < 0.05, except for temporal RNFL thickness). Global and regional cpCD relative loss values were higher than those of RNFL thickness (P < 0.05) in preperimetric glaucoma (except for the superonasal region) and glaucoma (except for the inferonasal and superonasal regions) eyes. Race, intraocular pressure (IOP), and cpCD were associated with greater cpCD than RNFL thickness loss in early glaucoma at the individual level (P < 0.05). Global measurements of capillary density (whole image capillary density and cpCD) had higher diagnostic accuracies than RNFL thickness in detecting preperimetric glaucoma and glaucoma (P < 0.05; except for cpCD/RNFL thickness comparison in glaucoma [P = 0.059]). Visual function was significantly associated with RNFL thickness and cpCD globally and in all regions (P < 0.05, except for temporal RNFL thickness-function association [P = 0.070]).

CONCLUSIONS

Associations between capillary density and visual function were found in the regions known to be at highest risk for damage in preperimetric glaucoma eyes and all regions of mild glaucoma eyes. In early glaucoma, capillary density loss was more pronounced than RNFL thickness loss. Individual characteristics influence the relative magnitudes of capillary density loss compared with RNFL thickness loss. Retinal nerve fiber layer thickness and microvascular assessments are complementary and yield valuable information for the detection of early damages seen in POAG.

Optical Coherence Tomography Evaluation of Peripapillary and Macular Structure Changes in Pre-perimetric Glaucoma, Early Perimetric Glaucoma, and Ocular Hypertension: A Systematic Review and Meta-Analysis

Background: This study aimed to assess the differences in the average and sectoral peripapillary retinal nerve fiber layer (pRNFL), macular ganglion cell plus inner plexiform layer (mGCIPL), and macular ganglion cell complex (mGCC) thickness using optical coherence tomography (OCT) in patients with pre-perimetric glaucoma (PPG) compared to those with early perimetric glaucoma (EG) and ocular hypertension (OHT).

Methods: A comprehensive literature search of the PubMed database, the Cochrane Library, and Embase was performed from inception to March 2021. The weighted mean difference (WMD) with the 95% confidence interval (CI) was pooled for continuous outcomes.

Results: Twenty-three cross-sectional studies comprising 2,574 eyes (1,101 PPG eyes, 1,233 EG eyes, and 240 OHT eyes) were included in the systematic review and meta-analysis. The pooled results demonstrated that the average pRNFL (WMD = 8.22, 95% CI = 6.32–10.12, P < 0.00001), mGCIPL (WMD = 4.83, 95% CI = 3.43–6.23, P < 0.00001), and mGCC (WMD = 7.19, 95% CI = 4.52–9.85, P < 0.00001) were significantly thinner in patients with EG than in those with PPG. The sectoral thickness of pRNFL, mGCIPL, and mGCC were also significantly lower in the EG eyes. In addition, the average pRNFL and mGCC were significantly thinner in the PPG eyes than those in the OHT eyes (pRNFL: WMD = −8.57, 95% CI = −9.88 to −7.27, P < 0.00001; mGCC: WMD = −3.23, 95% CI = −6.03 to −0.44, P = 0.02). Similarly, the sectoral pRNFL and mGCC were also significantly thinner in the PPG eyes than those in the OHT eyes.

Conclusion: OCT-based measurements of peripapillary and macular structural alterations can be used to distinguish PPG from EG and OHT, which can help understand the pathophysiology of glaucoma at earlier stages. Studies that employ clock hour classification methods and longitudinal studies are needed to verify our findings.

Systematic Review Registration:https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=239798 CRD42021239798

Comparison of spectral domain optical coherence tomography parameters between disc suspects and "pre-perimetric" glaucomatous discs classified on disc photo

Purpose:

The aim of this study was to compare SD-OCT parameters between disc suspects and “pre-perimetric” glaucomatous discs classified on disc photos.

Methods:

Disc photos of suspicious discs with normal Humphrey visual fields (HVF) were graded as normal or pre-perimetric glaucomatous based on the consensus of three masked glaucoma specialists. RNFL and GCL-IPL maps of SD-OCT (Cirrus OCT) of these eyes were studied. Quantitative RNFL parameters were compared. Both groups were also compared with respect to parameters being classified as abnormal (at the 1% level), and the pattern of GCL-IPL and NFL maps were assessed qualitatively and classified as normal or pre-perimetric glaucomatous by a masked glaucoma specialist.

Results:

The average and inferior RNFL thicknesses were decreased in pre-perimetric glaucomatous eyes compared to normal eyes (p 0.01) The average, minimal, inferotemporal and inferior sector GCL-IPL thicknesses were decreased in pre-perimetric glaucomatous eyes (all P < 0.002) The highest AUC was for the inferior RNFL thickness (0.771) followed by average RNFL thickness (0.757) The sensitivity and specificity for any one abnormal RNFL parameter was 71.9% and 59.7%, for GCL-IPL parameters was 70% and 69.1% The positive (PLR) and negative likelihood ratios (NLR) were 1.78 and 0.47 for RNFL and 2.26 and 0.43 for GCL-IPL parameters. For the qualitative assessment of RNFL and GCL-IPL maps, the sensitivity, specificity, PLR and NLR were 75%, 77.2%, 3.29, and 0.32, respectively.

Conclusion:

Pre-perimetric disc suspects had greater OCT changes compared to normal disc suspects. Qualitative assessment of RNFL and GCL-IPL maps had the highest discriminatory ability.

Predicting the central 10 degrees visual field in glaucoma by applying a deep learning algorithm to optical coherence tomography images

We aimed to develop a model to predict visual field (VF) in the central 10 degrees in patients with glaucoma, by training a convolutional neural network (CNN) with optical coherence tomography (OCT) images and adjusting the values with Humphrey Field Analyzer (HFA) 24–2 test. The training dataset included 558 eyes from 312 glaucoma patients and 90 eyes from 46 normal subjects. The testing dataset included 105 eyes from 72 glaucoma patients. All eyes were analyzed by the HFA 10-2 test and OCT; eyes in the testing dataset were additionally analyzed by the HFA 24-2 test. During CNN model training, the total deviation (TD) values of the HFA 10-2 test point were predicted from the combined OCT-measured macular retinal layers’ thicknesses. Then, the predicted TD values were corrected using the TD values of the innermost four points from the HFA 24-2 test. Mean absolute error derived from the CNN models ranged between 9.4 and 9.5 B. These values reduced to 5.5 dB on average, when the data were corrected using the HFA 24-2 test. In conclusion, HFA 10-2 test results can be predicted with a OCT images using a trained CNN model with adjustment using HFA 24-2 test.

Predicting the Glaucomatous Central 10-Degree Visual Field From Optical Coherence Tomography Using Deep Learning and Tensor Regression

PURPOSE

To predict the visual field (VF) of glaucoma patients within the central 10° from optical coherence tomography (OCT) measurements using deep learning and tensor regression.

DESIGN

Cross-sectional study.

METHODS

Humphrey 10-2 VFs and OCT measurements were carried out in 505 eyes of 304 glaucoma patients and 86 eyes of 43 normal subjects. VF sensitivity at each test point was predicted from OCT-measured thicknesses of macular ganglion cell layer + inner plexiform layer, retinal nerve fiber layer, and outer segment + retinal pigment epithelium. Two convolutional neural network (CNN) models were generated: (1) CNN-PR, which simply connects the output of the CNN to each VF test point; and (2) CNN-TR, which connects the output of the CNN to each VF test point using tensor regression. Prediction performance was assessed using 5-fold cross-validation through the root mean squared error (RMSE). For comparison, RMSE values were also calculated using multiple linear regression (MLR) and support vector regression (SVR). In addition, the absolute prediction error for predicting mean sensitivity in the whole VF was analyzed.

RESULTS

RMSE with the CNN-TR model averaged 6.32 ± 3.76 (mean ± standard deviation) dB. Significantly (P < .05) larger RMSEs were obtained with other models: CNN-PR (6.76 ± 3.86 dB), SVR (7.18 ± 3.87 dB), and MLR (8.56 ± 3.69 dB). The absolute mean prediction error for the whole VF was 2.72 ± 2.60 dB with the CNN-TR model.

CONCLUSION

The Humphrey 10-2 VF can be predicted from OCT-measured retinal layer thicknesses using deep learning and tensor regression.

Optical coherence tomography indices for diagnosis of chronic glaucoma in patients with diabetes mellitus: a pilot study

PURPOSE

To evaluate optical coherence tomography (OCT) parameters in patients with concomitant type-2 diabetes mellitus (DM) and primary open angle glaucoma (POAG) compared with patients with either of these diseases.

METHODS

Sixty eyes (52 patients) were divided into three groups. The first group included nonglaucomatous diabetic patients, the second included patients with POAG without DM, and the third included patients with both POAG and DM. Spectral domain OCT evaluation of the retinal nerve fiber layer (RNFL), ganglion cell complex (GCC), and optic disc parameters was performed. Visual field (VF) was measured for structural and functional correlation.

RESULTS

Significant differences were found in average RNFL, inferior RNFL, average GCC, inferior GCC, rim area, focal loss volume (FLV%), and global loss volume (GLV%) (P = 0.014, 0.001, 0.027, 0.006, 0.009, 0.043, and 0.001, respectively). The concomitant presence of DM and glaucoma was a risk factor for decreased average RNFL, inferior RNFL, rim area, and inferior GCC, and for increased GLV% (P = 0.034, 0.002, 0.014, 0.015, and 0.003, respectively). The inferior RNFL thickness had the largest significant area under the curve (P = 0.726; 90% sensitivity) at a specificity greater than 80% with a cutoff value of 105.38 μm (P = 0.005) compared with average RNFL, inferior GCC, rim area, and GLV% (P = 0.073, 0.25, 0.23, and 0.1, respectively). VF demonstrated the predominance of nasal scotomata in the diabetic group and arcuate scotoma in the glaucoma group (P < 0.001 and 0.03, respectively).

CONCLUSIONS

OCT could be a valuable tool for the detection and follow-up of POAG in diabetic patients. The inferior RNFL thickness could be a sensitive and a specific predictor for glaucoma diagnosis and progression in diabetic patients without retinopathy.

Diagnostic utility of neuroretinal rim thickness, measured in clock-hour sectors with HD optical coherence tomography, in preperimetric glaucoma

BACKGROUND

We evaluated the usefulness of neuroretinal rim (NRR) thicknesses, measured in clock-hour sectors with Cirrus HD optical coherence tomography, for diagnosing preperimetric glaucoma (PPG).

METHODS

This prospective study included 39 eyes of 39 patients with PPG and 39 eyes of 39 controls that were matched to patients for age and refractive error. We measured the circumpapillary retinal nerve fiber layer (cpRNFL) thickness, macular ganglion cell-inner plexiform layer (GCIPL) thickness, and optic nerve head (ONH) parameters with optical coherence tomography. The clock-hour NRR thicknesses were derived from a 360° circumferential rim thickness curve. We analyzed the area under the receiver operating characteristics curve (AUROC), cutoff values, and sensitivities at specificities of 90% and 95%.

RESULTS

The largest area under the receiver operating characteristics curves were observed for the NRR thickness at 6 o'clock (0.823), the inferior RNFL thickness (0.821), the average RNFL thickness (0.819), and the NRR thickness at 7 o'clock (0.818). The performance of the NRR thickness at 6 o'clock was comparable to the best performances of the cpRNFL, GCIPL, and ONH parameters (all p > 0.05).

CONCLUSION

The ability of the clock-hour NRR thickness assessment to diagnose PPG was comparable to the diagnostic abilities of cpRNFL, GCIPL, and ONH parameters. The best indicator of PPG was the NRR thickness parameter that was at 6 o'clock. This finding could play a role in detecting early structural changes in PPG.

Impact of optical coherence tomography on diagnostic decision-making by UK community optometrists: a clinical vignette study

Purpose

In recent years, there has been widespread investment in imaging technologies by community optometrists in the UK, most notably optical coherence tomography (OCT). The aim of the current study was to determine the value of OCT in the diagnosis of posterior segment diseases in a representative sample of community optometrists using a clinical vignette methodology.

Methods

A group of community optometrists (n = 50) initially completed a standardised training package on OCT interpretation followed by a computer‐based assessment featuring 52 clinical vignettes, containing images of healthy (n = 8) or glaucomatous (n = 18) discs or healthy (n = 8) or diseased (n = 18) fundi. Each vignette featured either a single fundus/disc photographic image, or a combination of a fundus/disc image with the corresponding OCT scan. An expert panel confirmed that the fundus images presented alone and those in combination with OCT data were of a similar level of difficulty and that the cases were typical of those seen in primary care. For each case, the optometrist selected their diagnosis from a pull‐down list and reported their confidence in their decision using a 10‐point Likert scale. Pairwise comparisons of the fundus image alone and fundus image/OCT combination were made for both diagnostic performance and confidence.

Results

The mean percentage of correct diagnoses using fundus imaging alone was 62% (95% CI 59–64%) and for the combination of fundus image/OCT was 80% (95% CI 77–82%). The mean false negative rate with fundus alone was 27% reducing to 13% with the OCT combination. Median confidence scores for fundus imaging alone was 8.0 (IQR 7.0–8.0) and 8.3 (IQR 8.0–9.0) for the combination. Improvements in performance and confidence were statistically significant (p < 0.001).

Conclusion

The results from this vignette study suggests that OCT improves optometrists’ diagnostic performance compared to fundus observation alone. These initial results suggest that OCT provides valuable additional data that could augment case‐finding for glaucoma and retinal disease; however, further research is needed to assess its diagnostic performance in a routine clinical practice setting.

Investigating the structure-function relationship using Goldmann V standard automated perimetry where glaucomatous damage is advanced

PURPOSE

To investigate if the structure-function relationship between circumpapillary retinal nerve fibre layer (cpRNFL) thickness and visual field (VF) thresholds is stronger when using the Goldmann V target rather than the Goldman III target where glaucomatous damage is advanced.

METHODS

Optical coherence tomography (OCT) and VF (Humphrey Field Analyzer 24-2 or 30-2) measurements with Goldmann III (SITA standard) and V (full-threshold) targets were carried out in 51 eyes of 51 patients with primary open angle glaucoma. The relationship between cpRNFL thicknesses in supero- and infero-temporal sectors, and VF sensitivity with the Goldmann III or V target was investigated.

RESULT

Visual field sensitivities (dB) both with the Goldmann III target and Goldmann V target showed a floor effect in the structure-function relationship against cpRNFL thickness, at approximately 60 μm. There was no significant relationship between visual field sensitivity measured with the Goldmann V target (dB scale: p = 0.12, 1/Lambert scale: p = 0.40; linear mixed models) and cpRNFL thickness, when corresponding visual field sensitivity, measured with the Goldmann III target, was <20 dB.

CONCLUSION

There was no improvement in the structure-function relationship using the Goldmann V target (full-threshold), compared to using the Goldmann III target (SITA standard), where glaucomatous damage was advanced.

Remote Grading of the Anterior Chamber Angle Using Goniophotographs and Optical Coherence Tomography: Implications for Telemedicine or Virtual Clinics

PURPOSE

To evaluate the agreement and accuracy of grading goniophotographs and anterior segment optical coherence tomography (AS-OCT) results for assessment of the anterior chamber angle, and elicit factors driving concordance between perceived grade and ground truth.

METHODS

Three clinicians evaluated the goniophotographs and AS-OCT results of 75 patients. Graders' impressions of the angle grade, trabecular pigmentation, and iris contour were compared with the ground truth gonioscopic examination result when physically performed by a senior optometrist. Percentage agreement and kappa statistics were calculated. Binary logistic regression was used to elicit factors for accurate grading.

RESULTS

Exact angle matches and binary (open or closed) evaluations were above guessing rate for all graders. There was a systematic bias toward underestimating the angle structure across all graders, especially at the superior angle, by approximately 1 ordinal unit. Kappa statistics showed fair-moderate agreement for exact (0.387-0.520) and binary (0.347-0.520) angle evaluations. Agreement was unchanged when using a multimodal approach (0.373-0.523). Factors driving concordance were primarily related to the extremes of the anterior chamber angle configuration (shallow or deep structures, and iris contour). However, prediction models did not fully explain the levels of concordance with the ground truth (maximum R 2 amongst models 0.177).

CONCLUSIONS

Although moderate agreement between graders and ground truth could be obtained under binary evaluations, angle grades were generally underestimated. Factors affecting concordance were primarily the extremes of the ground truth angle and iris contour.

TRANSLATIONAL RELEVANCE

We highlight factors affecting accuracy of grading goniophotography and AS-OCT images of the anterior chamber angle.

Optical coherence tomography for glaucoma diagnosis: An evidence based meta-analysis

Purpose

Early detection, monitoring and understanding of changes in the retina are central to the diagnosis of glaucomatous optic neuropathy, and vital to reduce visual loss from this progressive condition. The main objective of this investigation was to compare glaucoma diagnostic accuracy of commercially available optical coherence tomography (OCT) devices (Zeiss Stratus, Zeiss Cirrus, Heidelberg Spectralis and Optovue RTVue, and Topcon 3D-OCT).

Patients

16,104 glaucomatous and 11,543 normal eyes reported in 150 studies.

Methods

Between Jan. 2017 and Feb 2017, MEDLINE^®, EMBASE^®, CINAHL^®, Cochrane Library^®, Web of Science^®, and BIOSIS^® were searched for studies assessing glaucoma diagnostic accuracy of the aforementioned OCT devices. Meta-analysis was performed pooling area under the receiver operating characteristic curve (AUROC) estimates for all devices, stratified by OCT type (RNFL, macula), and area imaged.

Results

150 studies with 16,104 glaucomatous and 11,543 normal control eyes were included. Key findings: AUROC of glaucoma diagnosis for RNFL average for all glaucoma patients was 0.897 (0.887–0.906, n = 16,782 patient eyes), for macula ganglion cell complex (GCC) was 0.885 (0.869–0.901, n = 4841 eyes), for macula ganglion cell inner plexiform layer (GCIPL) was 0.858 (0.835–0.880, n = 4211 eyes), and for total macular thickness was 0.795 (0.754–0.834, n = 1063 eyes).

Conclusion

The classification capability was similar across all 5 OCT devices. More diagnostically favorable AUROCs were demonstrated in patients with increased glaucoma severity. Diagnostic accuracy of RNFL and segmented macular regions (GCIPL, GCC) scans were similar and higher than total macular thickness. This study provides a synthesis of contemporary evidence with features of robust inclusion criteria and large sample size. These findings may provide guidance to clinicians when navigating this rapidly evolving diagnostic area characterized by numerous options.

Evaluation of spectral domain optical coherence tomography parameters in ocular hypertension, preperimetric, and early glaucoma

Purpose:

The objective of this study is to evaluate the diagnostic ability of retinal nerve fiber layer (RNFL), macular, optic nerve head (ONH) parameters in healthy subjects, ocular hypertension (OHT), preperimetric glaucoma (PPG), and early glaucoma (EG) patients, to reveal factors affecting the diagnostic ability of spectral domain-optical coherence tomography (SD-OCT) parameters and risk factors for glaucoma.

Methods:

Three hundred and twenty-six eyes (89 healthy, 77 OHT, 94 PPG, and 66 EG eyes) were analyzed. RNFL, macular, and ONH parameters were measured with SD-OCT. The area under the receiver operating characteristic curve (AUC) and sensitivity at 95% specificity was calculated. Logistic regression analysis was used to determine the glaucoma risk factors. Receiver operating characteristic regression analysis was used to evaluate the influence of covariates on the diagnostic ability of parameters.

Results:

In PPG patients, parameters that had the largest AUC value were average RNFL thickness (0.83) and rim volume (0.83). In EG patients, parameter that had the largest AUC value was average RNFL thickness (0.98). The logistic regression analysis showed average RNFL thickness was a risk factor for both PPG and EG. Diagnostic ability of average RNFL and average ganglion cell complex thickness increased as disease severity increased. Signal strength index did not affect diagnostic abilities. Diagnostic ability of average RNFL and rim area increased as disc area increased.

Conclusion:

When evaluating patients with glaucoma, patients at risk for glaucoma, and healthy controls RNFL parameters deserve more attention in clinical practice. Further studies are needed to fully understand the influence of covariates on the diagnostic ability of OCT parameters.

Goldmann V Standard Automated Perimetry Underestimates Central Visual Sensitivity in Glaucomatous Eyes with Increased Axial Length

PURPOSE

To investigate the effect of axial length (AL) on the structure-function relationship between retinal nerve fiber layer (RNFL) thickness measurements and visual field (VF) sensitivity measured with Goldmann III and V.

METHOD

There were 85 eyes of 85 patients with primary open angle glaucoma included in the current study. Optical coherence tomography and VF (Humphrey Field Analyzer 24-2 or 30-2) measurements with Goldmann III and V targets were carried out in all patients. The optic disc and the VF were divided into six clusters and the relationship between circumpapillary RNFL (cpRNFL) thickness and VF sensitivity (with Goldmann III or V), age, and AL were investigated in each cluster.

RESULT

Visual sensitivity with Goldmann III (19.3 ± 11.7 dB, mean ± standard deviation) was significantly lower than that with Goldmann V (24.6 ± 11.0 dB, P < 0.001, linear mixed model). Visual sensitivities with both Goldmann III and V were significantly correlated with cpRNFL thickness in all clusters. Visual sensitivity decreased with increasing AL in the nasal retinal area for both targets, however, this phenomenon was only observed with the Goldmann V target in the temporal area.

CONCLUSION

Visual sensitivity measured with the size V target decreases with increasing AL in the temporal area, which corresponds to the papillomacular bundle. In the nasal retinal area, visual sensitivity decreases with the increase of AL for both Goldmann III and Goldmann V.

TRANSLATIONAL RELEVANCE

Careful consideration is needed when measuring visual sensitivity using Goldmann V target in glaucomatous eyes with increased AL.

Diagnostic accuracy of optical coherence tomography for diagnosing glaucoma: secondary analyses of the GATE study

BACKGROUND/AIMS

To assess the diagnostic performance of retinal nerve fibre layer (RNFL) data of optical coherence tomography (OCT) for detecting glaucoma.

METHODS

Secondary analyses of a prospective, multicentre diagnostic study (Glaucoma Automated Tests Evaluation (GATE)) referred to hospital eye services in the UK were conducted. We included data from 899 of 966 participants referred to hospital eye services with suspected glaucoma or ocular hypertension. We used both eyes' data and logistic regression-based receiver operator characteristics analysis to build a set of models to measure the sensitivity and specificity of the average and inferior quadrant RNFL thickness data of OCT. The reference standard was expert clinician examination including automated perimetry. The main outcome measures were sensitivity at 0.95 specificity and specificity at 0.95 sensitivity and the corresponding RNFL thickness thresholds. We explored the possibility of accuracy improvement by adding measures of within-eye and between-eye variation, scan quality, intraocular pressure (IOP) and age.

RESULTS

Glaucoma was diagnosed in at least one eye in 17% of participants. Areas under the curve were between 0.83 and 0.88. When specificity was fixed at 0.95, the sensitivity was between 0.38 and 0.55, and the highest values were reached with models including the inferior quadrant rather than the average RNFL thickness. Fixing sensitivity at 0.95, the specificity was between 0.36 and 0.58. The addition of age, refractive error, IOP or within-subject variation did not improve the accuracy.

CONCLUSION

RNFL thickness data of OCT can be used as a diagnostic test, but accuracy estimates remain moderate even in exploratory multivariable modelling of aiming to improve accuracy.

The association between structure-function relationships and cognitive impairment in elderly glaucoma patients

Accurate measurement of visual field (VF) is important in accessing glaucoma, however this may not be achieved in patients with dementia or mild cognitive impairment (CI). We investigated the association between CI and structure-function relationships in elderly glaucoma patients. The study included 94 eyes of 51 glaucoma patients aged ≥75 years with no diagnoses of dementia. CI was assessed using the Mini Mental State Examination (MMSE). Using the leave-one-out cross-validation, the mean deviation (MD) of the Humphrey 30-2 VF was predicted from measurements of optical coherence tomography, and the relationship between the squared prediction error and the MMSE score, together with age, fixation loss (FL), false positive (FP), and false negative (FN) percentages that were analyzed using the linear mixed model. A high prevalence of MCI or dementia was observed in the elderly population. The squared prediction error value of the MD was 17.0 ± 21.1 (mean ± standard deviation). The squared prediction error increased with decreasing MMSE total score, but age, FL, FP, and FN were not related. Careful consideration is needed when interpreting the VF results of these patients, because VF can be over- or underestimated, as suggested by the decreased structure-function relationships.

A comparison of the diagnostic ability of vessel density and structural measurements of optical coherence tomography in primary open angle glaucoma

Purpose

To compare the diagnostic abilities of vessel density measurements of the optic nerve head (ONH), peripapillary and macular regions on optical coherence tomography (OCT) angiography in eyes with primary open angle glaucoma (POAG) with that of the ONH rim area, peripapillary retinal nerve fiber layer (RNFL) thickness and the macular ganglion cell complex (GCC) thickness measurements.

Methods

In a cross sectional study, 78 eyes of 50 control subjects and 117 eyes of 67 POAG patients underwent vessel density and structural measurements with spectral domain OCT. POAG was diagnosed based on the masked evaluation of optic disc stereo photographs. Area under receiver operating characteristic curves (AUC) and sensitivities at fixed specificities of vessel densities in ONH, peripapillary and macular regions were compared with rim area, RNFL and GCC thickness.

Results

The AUC (sensitivity at 95% specificity) of average vessel densities within the ONH, peripapillary and macular region were 0.77 (31%), 0.85 (56%) and 0.70 (18%) respectively. The same of ONH rim area, average RNFL and GCC thickness were 0.94 (83%), 0.95 (72%) and 0.93 (62%) respectively. AUCs of vessel densities were significantly lower (p<0.05) than that of the corresponding structural measurements. Pre-treatment IOP (coefficient: 0.08) affected (p<0.05) the AUC of ONH vessel density but not of any other vessel density or structural measurements.

Conclusions

Diagnostic abilities of ONH, peripapillary and the macular vessel densities in POAG were significantly lower than ONH rim area, peripapillary RNFL and macular GCC measurements respectively. At fixed levels of glaucoma severity, the diagnostic ability of the ONH vessel density was significantly greater in eyes with higher pre-treatment IOP.

Validating the Usefulness of the "Random Forests" Classifier to Diagnose Early Glaucoma With Optical Coherence Tomography

PURPOSE

To validate the usefulness of the "Random Forests" classifier to diagnose early glaucoma with spectral-domain optical coherence tomography (SDOCT).

METHODS

design: Comparison of diagnostic algorithms.

SETTING

Multiple institutional practices.

STUDY PARTICIPANTS

Training dataset included 94 eyes of 94 open-angle glaucoma (OAG) patients and 84 eyes of 84 normal subjects and testing dataset included 114 eyes of 114 OAG patients and 82 eyes of 82 normal subjects. In both groups, OAG eyes with mean deviation (MD) values better than -5.0 dB were included.

OBSERVATION PROCEDURE

Using the training dataset, classifiers were built to discriminate between glaucoma and normal eyes using 84 OCT measurements using the Random Forests method, multiple logistic regression models based on backward or bidirectional stepwise model selection, a least absolute shrinkage and selection operator regression (LASSO) model, and a Ridge regression model.

MAIN OUTCOME MEASURES

Diagnostic accuracy.

RESULTS

With the testing data, the area under the receiver operating characteristic curve (AROC) with the Random Forests method (93.0%) was significantly (P < .05) larger than those with other models of the stepwise model selections (71.9%), LASSO model (89.6%), and Ridge model (89.2%).

CONCLUSION

It is useful to analyze multiple SDOCT parameters concurrently using the Random Forests method to diagnose glaucoma in early stages.

Diagnostic accuracy of posterior pole asymmetry analysis parameters of spectralis optical coherence tomography in detecting early unilateral glaucoma

Purpose:

To report the diagnostic ability of posterior pole asymmetry analysis (PPAA) parameters of spectralis optical coherence tomography (OCT) in detecting early unilateral glaucoma.

Methods:

A prospective, cross-sectional study which included 80 eyes of 80 normal subjects and 76 eyes of 76 patients with unilateral early primary open-angle glaucoma by Hodapp-Anderson-Parrish classification. All subjects were of age more than 18 years, best-corrected visual acuity 20/40 or better, and a refractive error within ± 5 diopter (D) sphere and ± 3 D cylinder. Control subjects had a normal ocular examination, intraocular pressure (IOP) <22 mmHg, no past history of high IOP, no family history of glaucoma, normal optic disc morphology, and visual field in both eyes. One eye of the control subject was randomly included. All eyes underwent OCT for retinal nerve fiber layer (RNFL) analysis and PPAA. The number of continuous black squares was noted in the asymmetry analysis (right-left + hemisphere asymmetry). The area under curve (AUC) was calculated for all OCT parameters.

Results:

The best value for AUC for RNFL analysis was 0.858 for the inferotemporal quadrant thickness. This was similar to the best value for AUC for PPAA which was 0.833 for the inferior macular thickness parameter (P = 0.5). The AUC for the right-left and the hemisphere asymmetry part of PPAA was 0.427 and 0.499, respectively.

Conclusion:

The macular thickness PPAA parameters were equally good as the RNFL parameters. However, the asymmetry analysis parameters performed poorly and need further refinement before its use in early unilateral glaucoma diagnosis.

Regional Comparisons of Optical Coherence Tomography Angiography Vessel Density in Primary Open-Angle Glaucoma

PURPOSE

To compare the diagnostic abilities of the vessel densities in optic nerve head (ONH), peripapillary, and macular regions measured using optical coherence tomography angiography (OCTA) in eyes with primary open-angle glaucoma (POAG), and to evaluate the effect of glaucoma severity (based on the mean deviation, MD), optic disc size, and pretreatment intraocular pressure (IOP).

DESIGN

Cross-sectional study.

METHODS

Seventy-eight eyes of 53 control subjects and 64 eyes of 39 POAG patients underwent OCTA imaging. Area under receiver operating characteristic (ROC) curves (AUC) and sensitivities at fixed specificities of vessel densities in ONH, peripapillary, and macular regions were analyzed. ROC regression was used to evaluate the effect of covariates on the diagnostic abilities.

RESULTS

The AUCs of ONH vessel densities ranged between 0.59 (superonasal sector) and 0.73 (average inside disc), peripapillary between 0.70 (nasal, superonasal and temporal) and 0.89 (inferotemporal), and macular between 0.56 (nasal) and 0.64 (temporal). AUC of the average peripapillary vessel density was significantly better than the average inside disc (P = .05) and macular (P = .005) measurement. MD showed a negative association with the AUCs of the vessel densities of all regions. Pretreatment IOP (coefficient: 0.09) showed a significant (P < .05) effect on the AUC of ONH vessel density.

CONCLUSIONS

Diagnostic ability of the vessel density parameters of OCTA was only moderate. Macular and inside disc densities had significantly lower diagnostic abilities in POAG than the peripapillary density. Diagnostic abilities of vessel densities increased with increasing severity of glaucoma and that of ONH vessel density with higher pretreatment IOPs.

Optic nerve head parameters of high-definition optical coherence tomography and Heidelberg retina tomogram in perimetric and preperimetric glaucoma

Background:

Heidelberg retina tomogram (HRT) and optical coherence tomography (OCT) are two widely used imaging modalities to evaluate the optic nerve head (ONH) in glaucoma.

Purpose:

To compare the ONH parameters of HRT3 and high-definition OCT (HD-OCT) and evaluate their diagnostic abilities in perimetric and preperimetric glaucoma.

Design:

Cross-sectional analysis.

Methods:

35 control eyes (24 subjects), 21 preperimetric glaucoma eyes (15 patients), and 64 perimetric glaucoma eyes (44 patients) from the Longitudinal Glaucoma Evaluation Study underwent HRT3 and HD-OCT examinations.

Statistical Analysis:

Agreement between the ONH parameters of HRT and HD-OCT were assessed using Bland-Altman plots. Diagnostic abilities of ONH parameters were evaluated using area under the receiver operating characteristic curves (AUCs), sensitivity at fixed specificity, and likelihood ratios (LR).

Results:

Optic disc area, vertical cup to disc ratio, and cup volume with HD-OCT were larger than with HRT, while the rim area was smaller with HD-OCT (P < 0.001 for all comparisons). AUCs of all HD-OCT ONH parameters (0.90-0.97 in perimetric and 0.62-0.71 in preperimetric glaucoma) were comparable (P > 0.10) to the corresponding HRT ONH parameters (0.81-0.95 in perimetric and 0.55-0.72 in preperimetric glaucoma). LRs associated with diagnostic categorization of ONH parameters of both HD-OCT and HRT were associated with larger effects on posttest probability of perimetric compared to preperimetric glaucoma.

Conclusions:

ONH measurements of HD-OCT and HRT3 cannot be used interchangeably. Though the diagnostic abilities of ONH parameters of HD-OCT and HRT in glaucoma were comparable, the same were significantly lower in preperimetric compared to perimetric glaucoma.

Macular Diagnostic Ability in OCT for Assessing Glaucoma in High Myopia

PURPOSE

To compare the diagnostic abilities of spectral-domain optical coherence tomography (SD-OCT; Spectralis OCT) and time-domain OCT (TD-OCT; Stratus OCT). Changes in macular parameters in highly myopic eyes of glaucoma patients and highly myopic eyes of glaucoma suspects were evaluated and compared.

METHODS

We collected data from 72 highly myopic eyes (spherical equivalent, ≤-6.0D). Forty-one eyes had perimetric glaucoma and 31 eyes were suspected to have glaucoma (control group). All eyes underwent SD-OCT and TD-OCT imaging. Area under the receiver operating characteristic (AUROC) curve and sensitivity were examined on macular volume and thickness parameters at a fixed specificity and compared between groups.

RESULTS

The highest TD-OCT AUROC curves were found using outer inferior sector macular thickness (AUROC curve, 0.911) and volume (AUROC curve, 0.909). The highest SD-OCT AUROC curves were found using outer inferior region thickness (AUROC curve, 0.836) and volume (AUROC curve, 0.834). The difference between the two imaging modalities was not statistically significant (thickness, p = 0.141; volume, p = 0.138). The sensitivity of TD-OCT macular outer inferior average thickness was highest and was 88.2%, with a specificity of 80.4%. The sensitivity of TD-OCT average volume measurements in this same region was 76.5%, with a specificity of 91.3%. The SD-OCT average thickness measurements also had the highest sensitivity in this region, which was 78.6%, with a specificity of 82.1%. The SD-OCT volume measurements had a sensitivity of 67.9%, with a specificity of 92.3%.

CONCLUSIONS

Both SD-OCT and TD-OCT measurements of outer inferior macular thickness and volume can differentiate between eyes of glaucoma patients and glaucoma suspects with high myopia. These independent predictors all had good sensitivity. Based on our results, SD-OCT and TD-OCT have similar diagnostic abilities. These parameters may provide useful additional data in highly myopic eyes to complement standard glaucoma diagnosis tools.

Decreased Glycemic Difference Between Diabetes and Nondiabetes in the Elderly Leads to the Reduced Diagnostic Accuracy of Hemoglobin A1c for Diabetes Screening in an Aged Chinese Population

BACKGROUND

This study investigated the impact of age on the accuracy of glycated hemoglobin (HbA1c) for diabetes screening and explored the possible cause(s).

MATERIALS AND METHODS

Data from 3,050 Chinese participants 25-75 years of age without known diabetes in a population-based cross-sectional survey were analyzed. Diabetes was diagnosed by the oral glucose tolerance test (OGTT). The performance of HbA1c for detecting OGTT-defined diabetes in tertile groups (divided by age) was evaluated by the area under the curve (AUC) of the receiver operating characteristic curve (ROC). The effect of age on the difference in glucose levels between participants with and without diabetes and the impact of this difference on the performance of HbA1c were evaluated.

RESULTS

In young (25-41 years old), middle-aged (41-53 years old), and old (55-72 years old) participants, the ROC AUC (95% confidence interval) of HbA1c for detecting OGTT-defined diabetes was 0.958 (0.915, 1.000), 0.891 (0.852, 0.930), and 0.861 (0.821, 0.901), respectively (P = 0.005). The difference of fasting plasma glucose between participants with diabetes and those without diabetes decreased with increasing age: 3.01 (2.80, 3.22) mmol/L, 2.90 (2.71, 3.09) mmol/L, and 2.33 (2.16, 2.50) mmol/L in the three consecutive age groups, respectively. A similar pattern was found in 2-h postprandial plasma glucose. The impact of age on the diagnostic power of HbA1c diminished after data were rearranged to artificially increase the difference between participants without diabetes and those with diabetes.

CONCLUSIONS

The accuracy of HbA1c for detecting OGTT-defined diabetes declines with age. This is largely due to the decreased separation in glycemic levels between participants with diabetes and without diabetes in the elderly.

Macular versus Retinal Nerve Fiber Layer Parameters for Diagnosing Manifest Glaucoma: A Systematic Review of Diagnostic Accuracy Studies

TOPIC

Macular parameters have been proposed as an alternative to retinal nerve fiber layer (RNFL) parameters to diagnose glaucoma. Comparing the diagnostic accuracy of macular parameters, specifically the ganglion cell complex (GCC) and ganglion cell inner plexiform layer (GCIPL), with the accuracy of RNFL parameters for detecting manifest glaucoma is important to guide clinical practice and future research.

METHODS

Studies using spectral domain optical coherence tomography (SD OCT) and reporting macular parameters were included if they allowed the extraction of accuracy data for diagnosing manifest glaucoma, as confirmed with automated perimetry or a clinician's optic nerve head (ONH) assessment. Cross-sectional cohort studies and case-control studies were included. The QUADAS 2 tool was used to assess methodological quality. Only direct comparisons of macular versus RNFL parameters (i.e., in the same study) were conducted. Summary sensitivity and specificity of each macular or RNFL parameter were reported, and the relative diagnostic odds ratio (DOR) was calculated in hierarchical summary receiver operating characteristic (HSROC) models to compare them.

RESULTS

Thirty-four studies investigated macular parameters using RTVue OCT (Optovue Inc., Fremont, CA) (19 studies, 3094 subjects), Cirrus OCT (Carl Zeiss Meditec Inc., Dublin, CA) (14 studies, 2164 subjects), or 3D Topcon OCT (Topcon, Inc., Tokyo, Japan) (4 studies, 522 subjects). Thirty-two of these studies allowed comparisons between macular and RNFL parameters. Studies generally reported sensitivities at fixed specificities, more commonly 0.90 or 0.95, with sensitivities of most best-performing parameters between 0.65 and 0.75. For all OCT devices, compared with RNFL parameters, macular parameters were similarly or slightly less accurate for detecting glaucoma at the highest reported specificity, which was confirmed in analyses at the lowest specificity. Included studies suffered from limitations, especially the case-control study design, which is known to overestimate accuracy. However, this flaw is less relevant as a source of bias in direct comparisons conducted within studies.

CONCLUSIONS

With the use of OCT, RNFL parameters are still preferable to macular parameters for diagnosing manifest glaucoma, but the differences are small. Because of high heterogeneity, direct comparative or randomized studies of OCT devices or OCT parameters and diagnostic strategies are essential.

Reference Standard Test and the Diagnostic Ability of Spectral Domain Optical Coherence Tomography in Glaucoma

PURPOSE

To evaluate the relationship between the reference standard used to diagnose glaucoma and the diagnostic ability of spectral domain optical coherence tomograph (SDOCT).

METHODS

In a cross-sectional study, 280 eyes of 175 consecutive subjects, referred to a tertiary eye care center for glaucoma evaluation, underwent optic disc photography, visual field (VF) examination, and SDOCT examination. The cohort was divided into glaucoma and control groups based on 3 reference standards for glaucoma diagnosis: first based on the optic disc classification (179 glaucoma and 101 control eyes), second on VF classification (glaucoma hemifield test outside normal limits and pattern SD with P-value of <5%, 130 glaucoma and 150 control eyes), and third on the presence of both glaucomatous optic disc and glaucomatous VF (125 glaucoma and 155 control eyes). Relationship between the reference standards and the diagnostic parameters of SDOCT were evaluated using areas under the receiver operating characteristic curve, sensitivity, and specificity.

RESULTS

Areas under the receiver operating characteristic curve and sensitivities of most of the SDOCT parameters obtained with the 3 reference standards (ranging from 0.74 to 0.88 and 72% to 88%, respectively) were comparable (P>0.05). However, specificities of SDOCT parameters were significantly greater (P<0.05) with optic disc classification as reference standard (74% to 88%) compared with VF classification as reference standard (57% to 74%).

CONCLUSIONS

Diagnostic parameters of SDOCT that was significantly affected by reference standard was the specificity, which was greater with optic disc classification as the reference standard. This has to be considered when comparing the diagnostic ability of SDOCT across studies.

Optic nerve head and fibre layer imaging for diagnosing glaucoma

BACKGROUND

The diagnosis of glaucoma is traditionally based on the finding of optic nerve head (ONH) damage assessed subjectively by ophthalmoscopy or photography or by corresponding damage to the visual field assessed by automated perimetry, or both. Diagnostic assessments are usually required when ophthalmologists or primary eye care professionals find elevated intraocular pressure (IOP) or a suspect appearance of the ONH. Imaging tests such as confocal scanning laser ophthalmoscopy (HRT), optical coherence tomography (OCT) and scanning laser polarimetry (SLP, as used by the GDx instrument), provide an objective measure of the structural changes of retinal nerve fibre layer (RNFL) thickness and ONH parameters occurring in glaucoma.

OBJECTIVES

To determine the diagnostic accuracy of HRT, OCT and GDx for diagnosing manifest glaucoma by detecting ONH and RNFL damage.

SEARCH METHODS

We searched several databases for this review. The most recent searches were on 19 February 2015.

SELECTION CRITERIA

We included prospective and retrospective cohort studies and case-control studies that evaluated the accuracy of OCT, HRT or the GDx for diagnosing glaucoma. We excluded population-based screening studies, since we planned to consider studies on self-referred people or participants in whom a risk factor for glaucoma had already been identified in primary care, such as elevated IOP or a family history of glaucoma. We only considered recent commercial versions of the tests: spectral domain OCT, HRT III and GDx VCC or ECC.

DATA COLLECTION AND ANALYSIS

We adopted standard Cochrane methods. We fitted a hierarchical summary ROC (HSROC) model using the METADAS macro in SAS software. After studies were selected, we decided to use 2 x 2 data at 0.95 specificity or closer in meta-analyses, since this was the most commonly-reported level.

MAIN RESULTS

We included 106 studies in this review, which analysed 16,260 eyes (8353 cases, 7907 controls) in total. Forty studies (5574 participants) assessed GDx, 18 studies (3550 participants) HRT, and 63 (9390 participants) OCT, with 12 of these studies comparing two or three tests. Regarding study quality, a case-control design in 103 studies raised concerns as it can overestimate accuracy and reduce the applicability of the results to daily practice. Twenty-four studies were sponsored by the manufacturer, and in 15 the potential conflict of interest was unclear.Comparisons made within each test were more reliable than those between tests, as they were mostly based on direct comparisons within each study.The Nerve Fibre Indicator yielded the highest accuracy (estimate, 95% confidence interval (CI)) among GDx parameters (sensitivity: 0.67, 0.55 to 0.77; specificity: 0.94, 0.92 to 0.95). For HRT measures, the Vertical Cup/Disc (C/D) ratio (sensitivity: 0.72, 0.60 to 0.68; specificity: 0.94, 0.92 to 0.95) was no different from other parameters. With OCT, the accuracy of average RNFL retinal thickness was similar to the inferior sector (0.72, 0.65 to 0.77; specificity: 0.93, 0.92 to 0.95) and, in different studies, to the vertical C/D ratio.Comparing the parameters with the highest diagnostic odds ratio (DOR) for each device in a single HSROC model, the performance of GDx, HRT and OCT was remarkably similar. At a sensitivity of 0.70 and a high specificity close to 0.95 as in most of these studies, in 1000 people referred by primary eye care, of whom 200 have manifest glaucoma, such as in those who have already undergone some functional or anatomic testing by optometrists, the best measures of GDx, HRT and OCT would miss about 60 cases out of the 200 patients with glaucoma, and would incorrectly refer 50 out of 800 patients without glaucoma. If prevalence were 5%, e.g. such as in people referred only because of family history of glaucoma, the corresponding figures would be 15 patients missed out of 50 with manifest glaucoma, avoiding referral of about 890 out of 950 non-glaucomatous people.Heterogeneity investigations found that sensitivity estimate was higher for studies with more severe glaucoma, expressed as worse average mean deviation (MD): 0.79 (0.74 to 0.83) for MD < -6 db versus 0.64 (0.60 to 0.69) for MD ≥ -6 db, at a similar summary specificity (0.93, 95% CI 0.92 to 0.94 and, respectively, 0.94; 95% CI 0.93 to 0.95; P < 0.0001 for the difference in relative DOR).

AUTHORS' CONCLUSIONS

The accuracy of imaging tests for detecting manifest glaucoma was variable across studies, but overall similar for different devices. Accuracy may have been overestimated due to the case-control design, which is a serious limitation of the current evidence base.We recommend that further diagnostic accuracy studies are carried out on patients selected consecutively at a defined step of the clinical pathway, providing a description of risk factors leading to referral and bearing in mind the consequences of false positives and false negatives in the setting in which the diagnostic question is made. Future research should report accuracy for each threshold of these continuous measures, or publish raw data.

Comparing spectral-domain optical coherence tomography and standard automated perimetry to diagnose glaucomatous optic neuropathy

PURPOSE

To compare the abilities of standard automated perimetry (SAP) and spectral-domain optical coherence tomography (SDOCT) in diagnosing eyes with glaucomatous optic neuropathy (GON).

METHODS

In a cross-sectional study, 280 eyes of 175 subjects referred to tertiary eye care center by general ophthalmologists for a glaucoma evaluation underwent retinal nerve fiber layer (RNFL) and ganglion cell complex (GCC) imaging with SDOCT. All subjects had at least 2 reliable and repeatable SAP. Two glaucoma experts masked to clinical and SAP results classified the optic nerves into GON and nonglaucomatous groups based on digital optic disc photographs. Ability of SDOCT parameters and SAP to discriminate GON eyes from nonglaucomatous eyes was evaluated using sensitivity, specificity, and likelihood ratios (LR).

RESULTS

Experts classified 179 eyes into GON and 101 eyes into nonglaucomatous group. Sensitivity of SAP (69.8%) was significantly lower (P<0.05) than that of inferior quadrant RNFL thickness (83.8%), average RNFL thickness (77.2%), GCC focal loss volume (FLV, 81.5%), and global loss volume (GLV, 82.6%). Specificity and positive LR of SAP (95% and14.1, respectively) were significantly greater than those of all RNFL parameters, FLV (84.2% and 5.1) and GLV (82.2% and 4.6). Negative LR of SAP (0.32) was significantly inferior to that of inferior quadrant RNFL thickness (0.22), FLV (0.22), and GLV (0.21).

CONCLUSIONS

Most of the RNFL and GCC parameters of SDOCT had better sensitivities and negative LRs to diagnose GON compared with SAP. The specificities and positive LRs of most SDOCT parameters were inferior to that of SAP.

The ISNT rule in glaucoma: revisiting with spectral domain optical coherence tomography

PURPOSE

To evaluate the ability of the neuroretinal rim (NRR) rules determined using spectral domain optical coherence tomography (SDOCT) in diagnosing glaucoma and to study the effect of optic disc size and disease severity on the diagnostic ability of these rules.

METHODS

In a cross-sectional study, 125 eyes of 96 glaucoma patients and 96 eyes of 72 control subjects underwent optic nerve head (ONH) imaging with SDOCT. Inferior (I), superior (S), nasal (N) and temporal (T) NRR areas were automatically determined by the sdoct software. Diagnostic abilities of ISNT (I > S > N > T), IT (I > T) and ST (S > T) rules in glaucoma were evaluated using sensitivity, specificity and likelihood ratios (LR). Effect of optic disc size and disease severity [based on mean deviation (MD) on visual fields] on the diagnostic ability of the NRR rules was evaluated using regression models.

RESULTS

Sensitivities of ISNT, IT and ST rules were 80.8%, 60.0% and 29.6%, respectively, and the specificities were 32.3%, 84.4% and 93.8%, respectively. Positive LRs of ISNT, IT and ST rules were 1.19, 3.84 and 4.74, respectively, and negative LRs were 0.60, 0.47 and 0.75, respectively. Sensitivities of ISNT (coefficient: -1.06, p = 0.02) and IT (-0.71, 0.05) rules decreased with increasing disc size. Positive LR of IT rule increased significantly (-0.01, 0.04) with decreasing MD, and negative LR of IT rule decreased (got better) significantly (0.26, 0.05) with decreasing disc size.

CONCLUSIONS

Neuroretinal rim rules, as determined by SDOCT, do not allow robust differentiation of glaucomatous from non-glaucomatous discs.

Atypical birefringence pattern and the diagnostic ability of scanning laser polarimetry with enhanced corneal compensation in glaucoma

PURPOSE

To evaluate the effect of typical scan score (TSS), when within the acceptable limits, on the diagnostic performance of retinal nerve fibre layer (RNFL) parameters with the enhanced corneal compensation (ECC) protocol of scanning laser polarimetry (SLP) in glaucoma.

METHODS

In a cross-sectional study, 203 eyes of 160 glaucoma patients and 140 eyes of 104 control subjects underwent RNFL imaging with the ECC protocol of SLP. TSS was used to quantify atypical birefringence pattern (ABP) images. Influence of TSS on the diagnostic ability of SLP parameters was evaluated by receiver operating characteristic (ROC) regression models after adjusting for the effect of disease severity [based on mean deviation (MD)] on standard automated perimetry).

RESULTS

Diagnostic abilities of all RNFL parameters of SLP increased when the TSS values were higher. This effect was statistically significant for TSNIT (coefficient: 0.08, p<0.001) and inferior average parameters (coefficient: 0.06, p=0.002) but not for nerve fibre indicator (NFI, coefficient: 0.03, p=0.21). In early glaucoma (MD of -5 dB), predicted area under ROC curve (AUC) for TSNIT average parameter improved from 0.642 at a TSS of 90 to 0.845 at a TSS of 100. In advanced glaucoma (MD of -15 dB), AUC for TSNIT average improved from 0.832 at a TSS of 90 to 0.947 at 100.

CONCLUSION

Diagnostic performances of TSNIT and inferior average RNFL parameters with ECC protocol of SLP were significantly influenced by TSS even when the TSS values were within the acceptable limits. Diagnostic ability of NFI was unaffected by TSS values.

Factors affecting the ability of the spectral domain optical coherence tomograph to detect photographic retinal nerve fiber layer defects

Purpose

To evaluate the ability of normative database classification (color-coded maps) of spectral domain optical coherence tomograph (SDOCT) in detecting wedge shaped retinal nerve fiber layer (RNFL) defects identified on photographs and the factors affecting the ability of SDOCT in detecting these RNFL defects.

Methods

In a cross-sectional study, 238 eyes (476 RNFL quadrants) of 172 normal subjects and 85 eyes (103 RNFL quadrants with wedge shaped RNFL defects) of 66 glaucoma patients underwent RNFL imaging with SDOCT. Logistic regression models were used to evaluate the factors associated with false positive and false negative RNFL classifications of the color-coded maps of SDOCT.

Results

False positive classification at a p value of <5% was seen in 108 of 476 quadrants (22.8%). False negative classification at a p value of <5% was seen in 16 of 103 quadrants (15.5%). Of the 103 quadrants with RNFL defects, 64 showed a corresponding VF defect in the opposite hemisphere and 39 were preperimetric. Higher signal strength index (SSI) of the scan was less likely to have a false positive classification (odds ratio: 0.97, p = 0.01). Presence of an associated visual field defect (odds ratio: 0.17, p = 0.01) and inferior quadrant RNFL defects as compared to superior (odds ratio: 0.24, p = 0.04) were less likely to show false negative classifications.

Conclusions

Scans with lower signal strengths were more likely to show false positive RNFL classifications, and preperimetric and superior quadrant RNFL defects were more likely to show false negative classifications on color-coded maps of SDOCT.

Peripapillary retinal nerve fiber layer assessment of spectral domain optical coherence tomography and scanning laser polarimetry to diagnose preperimetric glaucoma

Purpose

To compare the abilities of peripapillary retinal nerve fiber layer (RNFL) parameters of spectral domain optical coherence tomograph (SDOCT) and scanning laser polarimeter (GDx enhanced corneal compensation; ECC) in detecting preperimetric glaucoma.

Methods

In a cross-sectional study, 35 preperimetric glaucoma eyes (32 subjects) and 94 control eyes (74 subjects) underwent digital optic disc photography and RNFL imaging with SDOCT and GDx ECC. Ability of RNFL parameters of SDOCT and GDx ECC to discriminate preperimetric glaucoma eyes from control eyes was compared using area under receiver operating characteristic curves (AUC), sensitivities at fixed specificities and likelihood ratios (LR).

Results

AUC of the global average RNFL thickness of SDOCT (0.786) was significantly greater (p<0.001) than that of GDx ECC (0.627). Sensitivities at 95% specificity of the corresponding parameters were 20% and 8.6% respectively. AUCs of the inferior, superior and temporal quadrant RNFL thickness parameters of SDOCT were also significantly (p<0.05) greater than the respective RNFL parameters of GDx ECC. LRs of outside normal limits category of SDOCT parameters ranged between 3.3 and 4.0 while the same of GDx ECC parameters ranged between 1.2 and 2.1. LRs of within normal limits category of SDOCT parameters ranged between 0.4 and 0.7 while the same of GDx ECC parameters ranged between 0.7 and 1.0.

Conclusions

Abilities of the RNFL parameters of SDOCT and GDx ECC to diagnose preperimetric glaucoma were only moderate. Diagnostic abilities of the RNFL parameters of SDOCT were significantly better than that of GDx ECC in preperimetric glaucoma.

Discriminating between glaucoma and normal eyes using optical coherence tomography and the 'Random Forests' classifier

PURPOSE

To diagnose glaucoma based on spectral domain optical coherence tomography (SD-OCT) measurements using the 'Random Forests' method.

METHODS

SD-OCT was conducted in 126 eyes of 126 open angle glaucoma (OAG) patients and 84 eyes of 84 normal subjects. The Random Forests method was then applied to discriminate between glaucoma and normal eyes using 151 OCT parameters including thickness measurements of circumpapillary retinal nerve fiber layer (cpRNFL), the macular RNFL (mRNFL) and the ganglion cell layer-inner plexiform layer combined (GCIPL). The area under the receiver operating characteristic curve (AROC) was calculated using the Random Forests method adopting leave-one-out cross validation. For comparison, AROCs were calculated based on each one of the 151 OCT parameters.

RESULTS

The AROC obtained with the Random Forests method was 98.5% [95% Confidence interval (CI): 97.1-99.9%], which was significantly larger than the AROCs derived from any single OCT parameter (maxima were: 92.8 [CI: 89.4-96.2] %, 94.3 [CI: 91.1-97.6] % and 91.8 [CI: 88.2-95.4] % for cpRNFL-, mRNFL- and GCIPL-related parameters, respectively; P<0.05, DeLong's method with Holm's correction for multiple comparisons). The partial AROC above specificity of 80%, for the Random Forests method was equal to 18.5 [CI: 16.8-19.6] %, which was also significantly larger than the AROCs of any single OCT parameter (P<0.05, Bootstrap method with Holm's correction for multiple comparisons).

CONCLUSIONS

The Random Forests method, analyzing multiple SD-OCT parameters concurrently, significantly improves the diagnosis of glaucoma compared with using any single SD-OCT measurement.

Scanning the macula for detecting glaucoma

Background:

With the advent of spectral domain optical coherence tomography (SDOCT), there has been a renewed interest in macular region for detection of glaucoma. However, most macular SDOCT parameters currently are thickness parameters which evaluate thinning of the macular layers but do not quantify the extent of area over which the thinning has occurred. We therefore calculated a new macular parameter, ganglion cell complex surface abnormality ratio (GCC SAR) that represented the surface area over which the macular thickness was decreased.

Purpose:

To evaluate the ability of SAR in detecting perimetric and preperimetric glaucoma.

Design:

Retrospective image analysis.

Materials and Methods:

68 eyes with perimetric glaucoma, 62 eyes with preperimetric glaucoma and 165 control eyes underwent GCC imaging with SDOCT. SAR was calculated as the ratio of the abnormal to total area on the GCC significance map.

Statistical Analysis:

Diagnostic ability of SAR in glaucoma was compared against that of the standard parameters generated by the SDOCT software using area under receiver operating characteristic curves (AUC) and sensitivities at fixed specificities.

Results:

AUC of SAR (0.91) was statistically significantly better than that of GCC average thickness (0.86, P= 0.001) and GCC global loss volume (GLV; 0.88, P= 0.01) in differentiating perimetric glaucoma from control eyes. In differentiating preperimetric glaucoma from control eyes, AUC of SAR (0.72) was comparable to that of GCC average thickness (0.70, P> 0.05) and GLV (0.72, P> 0.05). Sensitivities at specificities of 80% and 95% of SAR were comparable (P > 0.05 for all comparisons) to that of GCC average thickness and GLV in diagnosing perimetric and preperimetric glaucoma.

Conclusion:

GCC SAR had a better ability to diagnose perimetric glaucoma compared to the SDOCT software provided global GCC parameters. However, in diagnosing preperimetric glaucoma, the ability of SAR was similar to that of software provided global GCC parameters.

The association of spectral-domain optical coherence tomography determined ganglion cell complex parameters and disease severity in Parkinson's disease

PURPOSE

To evaluate the association between the retinal structural changes provided by spectral domain optical coherence tomography (SD-OCT) and disease severity in Parkinson's disease (PD).

MATERIALS AND METHODS

Thirty-eight eyes of 20 patients with PD and 30 eyes of 30 healthy subjects were enrolled in this prospective study. The eyes ipsilateral (20 eyes) and contralateral (18 eyes) to the most affected body side in patients with PD were evaluated separately. SD-OCT (RTVue-100) was used to measure the macular ganglion cell complex thickness (mGCC), outer retinal thickness, and peripapillary retinal nerve fiber layer (RNFL) thickness. Disease severity was assessed using Unified Parkinson's Disease Rating Scale (UPDRS).

RESULTS

The differences in the nasal RNFL (p = 0.004), average mGCC (p = 0.014), superior mGCC (p = 0.007), inferior mGCC (p = 0.03) were significant between the ipsilateral eye group and controls. The differences in the nasal RNFL (p = 0.002), and superior mGCC (p = 0.038) were statisticallly significant between the contralateral eye group and controls. Inferior and average mGCC thicknesses of ipsilateral eyes were inversely correlated with UPDRS score (p = 0.001, r = -0.678 and p = 0.009, r = -0.568, respectively). Significant inverse correlation was demonstrated between UPDRS score and inferior RNFL of both eyes ipsilateral and contralateral to the most affected body side.

CONCLUSIONS

There is significant thinning in mGCC parameters of the retina in PD. The SD-OCT derived mGCC parameters may improve the detection of PD progression.

Retinal nerve fiber layer evaluation of spectral domain optical coherence tomograph and scanning laser polarimeter to diagnose glaucoma

PURPOSE

To compare the abilities of retinal nerve fiber layer (RNFL) parameters of spectral domain optical coherence tomograph (SDOCT) and scanning laser polarimeter (GDx enhanced corneal compensation; ECC) in detecting glaucoma.

METHODS

In a cross-sectional study, 215 eyes of 165 subjects (106 eyes of 79 glaucoma patients and 109 eyes of 86 controls) referred by general ophthalmologists for glaucoma evaluation underwent RNFL imaging with SDOCT and GDx ECC. Ability of RNFL parameters of SDOCT to discriminate glaucoma eyes from control eyes was compared with that of GDx ECC using area under operating characteristic curves (AUCs), sensitivities at fixed specificities, and likelihood ratios (LRs).

RESULTS

AUC of the average RNFL thickness of SDOCT to differentiate glaucoma from control eyes (0.868) was comparable (P=0.32) to that of GDx ECC (0.855). Sensitivity at 95% specificity was 63.2% for average RNFL thickness of SDOCT and 48.1% for the average RNFL measurement of GDx ECC. LRs of outside normal limits category of SDOCT parameters ranged between 5.6 and 7.7 while the same of GDx ECC parameters ranged between 3.1 and 3.7. LRs of within normal limits category of SDOCT parameters ranged between 0.18 and 0.24 while the same of GDx ECC parameters ranged between 0.20 and 0.32.

CONCLUSION

Though AUCs and sensitivities at fixed specificities were comparable between the RNFL parameters of SDOCT and GDx ECC in diagnosing glaucoma, LRs indicated that the RNFL parameters of SDOCT were better in 'ruling in' glaucoma.

Effect of scan quality on diagnostic accuracy of spectral-domain optical coherence tomography in glaucoma

PURPOSE

To evaluate the effect of scan quality on the diagnostic accuracies of optic nerve head (ONH), retinal nerve fiber layer (RNFL), and ganglion cell complex (GCC) parameters of spectral-domain optical coherence tomography (SD OCT) in glaucoma.

DESIGN

Cross-sectional study.

METHODS

Two hundred fifty-two eyes of 183 control subjects (mean deviation [MD]: -1.84 dB) and 207 eyes of 159 glaucoma patients (MD: -7.31 dB) underwent ONH, RNFL, and GCC scanning with SD OCT. Scan quality of SD OCT images was based on signal strength index (SSI) values. Influence of SSI on diagnostic accuracy of SD OCT was evaluated by receiver operating characteristic (ROC) regression.

RESULTS

Diagnostic accuracies of all SD OCT parameters were better when the SSI values were higher. This effect was statistically significant (P < .05) for ONH and RNFL but not for GCC parameters. In mild glaucoma (MD of -5 dB), area under ROC curve (AUC) for rim area, average RNFL thickness, and average GCC thickness parameters improved from 0.651, 0.678, and 0.726, respectively, at an SSI value of 30 to 0.873, 0.962, and 0.886, respectively, at an SSI of 70. AUCs of the same parameters in advanced glaucoma (MD of -15 dB) improved from 0.747, 0.890, and 0.873, respectively, at an SSI value of 30 to 0.922, 0.994, and 0.959, respectively, at an SSI of 70.

CONCLUSION

Diagnostic accuracies of SD OCT parameters in glaucoma were significantly influenced by the scan quality even when the SSI values were within the manufacturer-recommended limits. These results should be considered while interpreting the SD OCT scans for glaucoma.