ABSTRACT
BACKGROUND: Deep learning (DL) is promising to detect glaucoma. However, patients' privacy and data security are major concerns when pooling all data for model development. We developed a privacy-preserving DL model using the federated learning (FL) paradigm to detect glaucoma from optical coherence tomography (OCT) images. METHODS: This is a multicentre study. The FL paradigm consisted of a 'central server' and seven eye centres in Hong Kong, the USA and Singapore. Each centre first trained a model locally with its own OCT optic disc volumetric dataset and then uploaded its model parameters to the central server. The central server used FedProx algorithm to aggregate all centres' model parameters. Subsequently, the aggregated parameters are redistributed to each centre for its local model optimisation. We experimented with three three-dimensional (3D) networks to evaluate the stabilities of the FL paradigm. Lastly, we tested the FL model on two prospectively collected unseen datasets. RESULTS: We used 9326 volumetric OCT scans from 2785 subjects. The FL model performed consistently well with different networks in 7 centres (accuracies 78.3%-98.5%, 75.9%-97.0%, and 78.3%-97.5%, respectively) and stably in the 2 unseen datasets (accuracies 84.8%-87.7%, 81.3%-84.8%, and 86.0%-87.8%, respectively). The FL model achieved non-inferior performance in classifying glaucoma compared with the traditional model and significantly outperformed the individual models. CONCLUSION: The 3D FL model could leverage all the datasets and achieve generalisable performance, without data exchange across centres. This study demonstrated an OCT-based FL paradigm for glaucoma identification with ensured patient privacy and data security, charting another course toward the real-world transition of artificial intelligence in ophthalmology.
ABSTRACT
PURPOSE: The primary purpose of this review is to provide a comprehensive summary on the technical principles of OCTA and to enumerate vascular parameters being explicated for glaucoma diagnosis and progression with emphasis on recent studies. In addition, the authors also summarize the future clinical potentials of OCTA in glaucoma and enumerate the limitations of this imaging modality in the present-day scenario. METHODS: The index study is a narrative review on OCTA in glaucoma. The authors searched the PubMed database using the key phrases ''optical coherence tomography angiography" AND "glaucoma,'' AND/OR "vascular parameters" AND/OR "ocular perfusion." Being a relatively recent development in ocular imaging, studies in which OCTA imaging had been used for glaucoma evaluation since 2012 were included until March 2022. The literature search included original studies and previous review articles, while case reports were excluded. Preliminary search was based on relevant articles with search keywords in the title and abstract. The second screening was performed by reading the full text of the literature. RESULTS: Recent studies indicate reduction in microcirculation in glaucomatous eyes compared to the normal subjects. The area of interest for glaucoma evaluation using OCTA varies among the different studies. Based on the literature reviewed here, (1) OCTA parameters measured in the peripapillary; ONH and macular area have been shown to differentiate between glaucoma and normal eyes with a discriminatory power comparable to OCT parameters used routinely in clinics, (2) monitoring of peripapillary and macular vessel density may provide important information to the evaluation of glaucoma progression and prediction of rates of disease worsening, (3) studies suggest strong correlation between the OCTA parameters, the OCT parameters and visual function, measured by visual field testing, in glaucomatous eyes, (4) future prospects of OCTA in glaucoma evaluations using AI predicting structural and functional features and prognosis based on early vascular findings would open up scope for early detection of high-risk suspects and fast progressors in glaucoma. CONCLUSION: OCTA can be useful in quantifying vascular parameters in the optic disc, peripapillary and the macular regions for glaucoma evaluation. OCTA shows potential to become a part of everyday glaucoma management.
Subject(s)
Glaucoma , Tomography, Optical Coherence , Humans , Tomography, Optical Coherence/methods , Fluorescein Angiography/methods , Retinal Vessels/diagnostic imaging , Intraocular Pressure , Glaucoma/diagnosisABSTRACT
PURPOSE: Portable perimetric testing could be useful for community-based glaucoma screening programs. Frequency-doubling technology (FDT) and the Moorfields motion displacement test (MDT) are portable perimeters that have shown promise as potential screening tools for glaucoma. This study's goal was to determine the diagnostic accuracy of FDT and MDT for visual field defects and glaucoma. DESIGN: Prospective, cross-sectional, diagnostic accuracy study. PARTICIPANTS: A consecutive series of patients aged ≥ 50 years who presented to a glaucoma clinic in South India and had never undergone Humphrey field analyzer (HFA) visual field testing in the past. METHODS: Participants underwent 24-2 Swedish Interactive Thresholding Algorithm (SITA) Standard HFA perimetry, FDT perimetry, MDT perimetry, and iPad perimetry using visualFields Easy in random order. Ophthalmologist grades of HFA and optic nerve head photographs were used as reference standards for glaucoma and field defect presence. Receiver operating characteristic curves were constructed to assess the diagnostic accuracy of various parameters for each test. MAIN OUTCOME MEASURES: Sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). RESULTS: Overall, 292 eyes from 173 participants were included, with 112 eyes classified as moderate or worse glaucoma. For moderate or worse glaucoma detection, the best parameter on FDT was mean deviation (MD) (AUROC, 0.84; 95% confidence interval [CI], 0.79-0.89) and the best parameter on MDT was global probability of true damage (GPTD) (AUROC, 0.87; 95% CI, 0.82-0.91). When specificity was set to 90%, the sensitivity for detection of moderate or worse glaucoma was 55% (95% CI, 39%-68%) for FDT MD and 62% (95% CI 52%-71%) for MDT GPTD. CONCLUSIONS: Frequency-doubling technology and MDT perimetry had fair diagnostic accuracy for glaucoma detection when administered to naïve test takers in this South Indian population. Although not appropriate for use as a sole glaucoma screening test, these perimetric tests may be useful as ancillary tests. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.
Subject(s)
Glaucoma , Visual Fields , Humans , Cross-Sectional Studies , Prospective Studies , Sensitivity and Specificity , Glaucoma/diagnosis , Visual Field TestsABSTRACT
BACKGROUND/OBJECTIVES: Tablet-based perimetry could be used to test for glaucomatous visual field defects in settings without easy access to perimeters, although few studies have assessed diagnostic accuracy of tablet-based tests. The goal of this study was to determine the diagnostic accuracy of iPad perimetry using the visualFields Easy application. SUBJECTS/METHODS: This was a prospective, cross-sectional study of patients undergoing their first Humphrey Field Analyser (HFA) visual field test at a glaucoma clinic in India. Participants underwent 24-2 SITA Standard HFA testing and iPad-based perimetry with the visualFields Easy application. Reference standards for both visual field loss and suspected glaucoma were determined by ophthalmologist review of HFA results and optic disc photographs. Receiver operating characteristic curves were constructed to assess diagnostic accuracy at various test thresholds. RESULTS: 203 eyes from 115 participants were included, with 82 eyes classified as moderate or worse glaucoma. iPad perimetry had an area under the receiver operating characteristic (AUROC) curve of 0.64 (95% CI 0.57 to 0.71) for detection of any visual field defect relative to HFA and an AUROC of 0.68 (0.59 to 0.76) for detection of moderate or worse glaucoma relative to ophthalmologist examination. At a set specificity of 90%, the sensitivity of iPad perimetry for detection of moderate or worse glaucoma was 35% (22-48%). CONCLUSIONS: iPad perimetry using the visualFields Easy application had inadequate diagnostic accuracy to be used as a screening tool for glaucoma in this South Indian population.
Subject(s)
Glaucoma , Visual Field Tests , Humans , Visual Field Tests/methods , Visual Fields , Cross-Sectional Studies , Prospective Studies , Sensitivity and Specificity , Glaucoma/diagnosis , Glaucoma/epidemiology , ROC Curve , Vision Disorders/diagnosisABSTRACT
Purpose: We aim to develop a multi-task three-dimensional (3D) deep learning (DL) model to detect glaucomatous optic neuropathy (GON) and myopic features (MF) simultaneously from spectral-domain optical coherence tomography (SDOCT) volumetric scans. Methods: Each volumetric scan was labelled as GON according to the criteria of retinal nerve fibre layer (RNFL) thinning, with a structural defect that correlated in position with the visual field defect (i.e., reference standard). MF were graded by the SDOCT en face images, defined as presence of peripapillary atrophy (PPA), optic disc tilting, or fundus tessellation. The multi-task DL model was developed by ResNet with output of Yes/No GON and Yes/No MF. SDOCT scans were collected in a tertiary eye hospital (Hong Kong SAR, China) for training (80%), tuning (10%), and internal validation (10%). External testing was performed on five independent datasets from eye centres in Hong Kong, the United States, and Singapore, respectively. For GON detection, we compared the model to the average RNFL thickness measurement generated from the SDOCT device. To investigate whether MF can affect the model's performance on GON detection, we conducted subgroup analyses in groups stratified by Yes/No MF. The area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and accuracy were reported. Results: A total of 8,151 SDOCT volumetric scans from 3,609 eyes were collected. For detecting GON, in the internal validation, the proposed 3D model had significantly higher AUROC (0.949 vs. 0.913, p < 0.001) than average RNFL thickness in discriminating GON from normal. In the external testing, the two approaches had comparable performance. In the subgroup analysis, the multi-task DL model performed significantly better in the group of "no MF" (0.883 vs. 0.965, p-value < 0.001) in one external testing dataset, but no significant difference in internal validation and other external testing datasets. The multi-task DL model's performance to detect MF was also generalizable in all datasets, with the AUROC values ranging from 0.855 to 0.896. Conclusion: The proposed multi-task 3D DL model demonstrated high generalizability in all the datasets and the presence of MF did not affect the accuracy of GON detection generally.
ABSTRACT
Purpose: To develop a three-dimensional (3D) deep learning algorithm to detect glaucoma using spectral-domain optical coherence tomography (SD-OCT) optic nerve head (ONH) cube scans and validate its performance on ethnically diverse real-world datasets and on cropped ONH scans. Methods: In total, 2461 Cirrus SD-OCT ONH scans of 1012 eyes were obtained from the Glaucoma Clinic Imaging Database at the Byers Eye Institute, Stanford University, from March 2010 to December 2017. A 3D deep neural network was trained and tested on this unique raw OCT cube dataset to identify a multimodal definition of glaucoma excluding other concomitant retinal disease and optic neuropathies. A total of 1022 scans of 363 glaucomatous eyes (207 patients) and 542 scans of 291 normal eyes (167 patients) from Stanford were included in training, and 142 scans of 48 glaucomatous eyes (27 patients) and 61 scans of 39 normal eyes (23 patients) were included in the validation set. A total of 3371 scans (Cirrus SD-OCT) from four different countries were used for evaluation of the model: the non overlapping test dataset from Stanford (USA) consisted of 694 scans: 241 scans from 113 normal eyes of 66 patients and 453 scans of 157 glaucomatous eyes of 89 patients. The datasets from Hong Kong (total of 1625 scans; 666 OCT scans from 196 normal eyes of 99 patients and 959 scans of 277 glaucomatous eyes of 155 patients), India (total of 672 scans; 211 scans from 147 normal eyes of 98 patients and 461 scans from 171 glaucomatous eyes of 101 patients), and Nepal (total of 380 scans; 158 scans from 143 normal eyes of 89 patients and 222 scans from 174 glaucomatous eyes of 109 patients) were used for external evaluation. The performance of the model was then evaluated on manually cropped scans from Stanford using a new algorithm called DiagFind. The ONH region was cropped by identifying the appropriate zone of the image in the expected location relative to Bruch's Membrane Opening (BMO) using a commercially available imaging software. Subgroup analyses were performed in groups stratified by eyes, myopia severity of glaucoma, and on a set of glaucoma cases without field defects. Saliency maps were generated to highlight the areas the model used to make a prediction. The model's performance was compared to that of a glaucoma specialist using all available information on a subset of cases. Results: The 3D deep learning system achieved area under the curve (AUC) values of 0.91 (95% CI, 0.90-0.92), 0.80 (95% CI, 0.78-0.82), 0.94 (95% CI, 0.93-0.96), and 0.87 (95% CI, 0.85-0.90) on Stanford, Hong Kong, India, and Nepal datasets, respectively, to detect perimetric glaucoma and AUC values of 0.99 (95% CI, 0.97-1.00), 0.96 (95% CI, 0.93-1.00), and 0.92 (95% CI, 0.89-0.95) on severe, moderate, and mild myopia cases, respectively, and an AUC of 0.77 on cropped scans. The model achieved an AUC value of 0.92 (95% CI, 0.90-0.93) versus that of the human grader with an AUC value of 0.91 on the same subset of scans (\(P=0.99\)). The performance of the model in terms of recall on glaucoma cases without field defects was found to be 0.76 (0.68-0.85). Saliency maps highlighted the lamina cribrosa in glaucomatous eyes versus superficial retina in normal eyes as the regions associated with classification. Conclusions: A 3D convolutional neural network (CNN) trained on SD-OCT ONH cubes can distinguish glaucoma from normal cases in diverse datasets obtained from four different countries. The model trained on additional random cropping data augmentation performed reasonably on manually cropped scans, indicating the importance of lamina cribrosa in glaucoma detection. Translational Relevance: A 3D CNN trained on SD-OCT ONH cubes was developed to detect glaucoma in diverse datasets obtained from four different countries and on cropped scans. The model identified lamina cribrosa as the region associated with glaucoma detection.
Subject(s)
Deep Learning , Glaucoma , Myopia , Optic Disk , Optic Nerve Diseases , Glaucoma/diagnosis , Humans , Optic Disk/diagnostic imaging , Optic Nerve Diseases/diagnosisABSTRACT
PURPOSE: To determine the diagnostic accuracy of potential screening tests for moderate to advanced glaucoma. DESIGN: Prospective diagnostic test accuracy study. PARTICIPANTS: The study enrolled a consecutive series of patients aged ≥50 years who presented to a glaucoma clinic in South India without ever having received automated visual field testing. METHODS: All participants underwent 8 index tests: OCT of the peripapillary retinal nerve fiber layer, optic disc photography, Moorfield's Motion Displacement Test (MDT), frequency doubling technique perimetry, noncontact tonometry, pneumatonometry, presenting visual acuity, and best-corrected visual acuity. Participants also underwent stereoscopic photographs and Humphrey visual fields, which were used by 2 ophthalmologists to arrive at the reference standard diagnosis of moderate to advanced glaucoma. MAIN OUTCOME MEASURES: Sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio. RESULTS: A total of 217 people were enrolled; 321 eyes from 180 participants had all tests performed. Of these, 127 eyes (40%) were classified as having moderate to advanced glaucoma. Among the 8 tests, OCT best optimized sensitivity (84%, 95% confidence interval [CI], 76-90) and specificity (75%, 95% CI, 68-81). Moorfield's Motion Displacement Test was the best perimetric test, with a sensitivity of 91% (95% CI, 85-96) and specificity of 53% (95% CI, 44-61). Pressure and vision tests were not sensitive (e.g., sensitivity of 16%, 95% CI, 9-23 for noncontact tonometry and 23%, 95% CI, 15-31 for best-corrected visual acuity). Moorfield's Motion Displacement Test identified 16 of 127 eyes (13%) with glaucoma that were not captured by OCT, but also had false-positive results in 65 of 194 eyes (34%) without glaucoma that OCT correctly classified as negative. CONCLUSIONS: OCT had moderate sensitivity and fair specificity for diagnosing moderate to advanced glaucoma and should be prioritized during an initial assessment for glaucoma.
Subject(s)
Glaucoma , Nerve Fibers , Glaucoma/diagnosis , Humans , Manometry , Prospective Studies , Sensitivity and Specificity , Tomography, Optical Coherence/methods , Visual Acuity , Visual Field Tests/methodsABSTRACT
OBJECTIVE: Diabetic macular edema (DME) is the primary cause of vision loss among individuals with diabetes mellitus (DM). We developed, validated, and tested a deep learning (DL) system for classifying DME using images from three common commercially available optical coherence tomography (OCT) devices. RESEARCH DESIGN AND METHODS: We trained and validated two versions of a multitask convolution neural network (CNN) to classify DME (center-involved DME [CI-DME], non-CI-DME, or absence of DME) using three-dimensional (3D) volume scans and 2D B-scans, respectively. For both 3D and 2D CNNs, we used the residual network (ResNet) as the backbone. For the 3D CNN, we used a 3D version of ResNet-34 with the last fully connected layer removed as the feature extraction module. A total of 73,746 OCT images were used for training and primary validation. External testing was performed using 26,981 images across seven independent data sets from Singapore, Hong Kong, the U.S., China, and Australia. RESULTS: In classifying the presence or absence of DME, the DL system achieved area under the receiver operating characteristic curves (AUROCs) of 0.937 (95% CI 0.920-0.954), 0.958 (0.930-0.977), and 0.965 (0.948-0.977) for the primary data set obtained from CIRRUS, SPECTRALIS, and Triton OCTs, respectively, in addition to AUROCs >0.906 for the external data sets. For further classification of the CI-DME and non-CI-DME subgroups, the AUROCs were 0.968 (0.940-0.995), 0.951 (0.898-0.982), and 0.975 (0.947-0.991) for the primary data set and >0.894 for the external data sets. CONCLUSIONS: We demonstrated excellent performance with a DL system for the automated classification of DME, highlighting its potential as a promising second-line screening tool for patients with DM, which may potentially create a more effective triaging mechanism to eye clinics.
Subject(s)
Deep Learning , Diabetes Mellitus , Diabetic Retinopathy , Macular Edema , Diabetic Retinopathy/diagnostic imaging , Humans , Macular Edema/diagnostic imaging , ROC Curve , Tomography, Optical CoherenceABSTRACT
Purpose: The purpose of this study was to develop a 3D deep learning system from spectral domain optical coherence tomography (SD-OCT) macular cubes to differentiate between referable and nonreferable cases for glaucoma applied to real-world datasets to understand how this would affect the performance. Methods: There were 2805 Cirrus optical coherence tomography (OCT) macula volumes (Macula protocol 512 × 128) of 1095 eyes from 586 patients at a single site that were used to train a fully 3D convolutional neural network (CNN). Referable glaucoma included true glaucoma, pre-perimetric glaucoma, and high-risk suspects, based on qualitative fundus photographs, visual fields, OCT reports, and clinical examinations, including intraocular pressure (IOP) and treatment history as the binary (two class) ground truth. The curated real-world dataset did not include eyes with retinal disease or nonglaucomatous optic neuropathies. The cubes were first homogenized using layer segmentation with the Orion Software (Voxeleron) to achieve standardization. The algorithm was tested on two separate external validation sets from different glaucoma studies, comprised of Cirrus macular cube scans of 505 and 336 eyes, respectively. Results: The area under the receiver operating characteristic (AUROC) curve for the development dataset for distinguishing referable glaucoma was 0.88 for our CNN using homogenization, 0.82 without homogenization, and 0.81 for a CNN architecture from the existing literature. For the external validation datasets, which had different glaucoma definitions, the AUCs were 0.78 and 0.95, respectively. The performance of the model across myopia severity distribution has been assessed in the dataset from the United States and was found to have an AUC of 0.85, 0.92, and 0.95 in the severe, moderate, and mild myopia, respectively. Conclusions: A 3D deep learning algorithm trained on macular OCT volumes without retinal disease to detect referable glaucoma performs better with retinal segmentation preprocessing and performs reasonably well across all levels of myopia. Translational Relevance: Interpretation of OCT macula volumes based on normative data color distributions is highly influenced by population demographics and characteristics, such as refractive error, as well as the size of the normative database. Referable glaucoma, in this study, was chosen to include cases that should be seen by a specialist. This study is unique because it uses multimodal patient data for the glaucoma definition, and includes all severities of myopia as well as validates the algorithm with international data to understand generalizability potential.
Subject(s)
Deep Learning , Glaucoma , Macula Lutea , Optic Nerve Diseases , Glaucoma/diagnosis , Humans , Macula Lutea/diagnostic imaging , Tomography, Optical CoherenceABSTRACT
Glaucoma is the leading cause of irreversible blindness in the world. Structure and function assessments play an important role in diagnosing glaucoma. Nowadays, Optical Coherence Tomography (OCT) imaging gains increasing popularity in measuring the structural change of eyes. However, few automated methods have been developed based on OCT images to screen glaucoma. In this paper, we are the first to unify the structure analysis and function regression to distinguish glaucoma patients from normal controls effectively. Specifically, our method works in two steps: a semi-supervised learning strategy with smoothness assumption is first applied for the surrogate assignment of missing function regression labels. Subsequently, the proposed multi-task learning network is capable of exploring the structure and function relationship between the OCT image and visual field measurement simultaneously, which contributes to classification performance improvement. It is also worth noting that the proposed method is assessed by two large-scale multi-center datasets. In other words, we first build the largest glaucoma OCT image dataset (i.e., HK dataset) involving 975,400 B-scans from 4,877 volumes to develop and evaluate the proposed method, then the model without further fine-tuning is directly applied on another independent dataset (i.e., Stanford dataset) containing 246,200 B-scans from 1,231 volumes. Extensive experiments are conducted to assess the contribution of each component within our framework. The proposed method outperforms the baseline methods and two glaucoma experts by a large margin, achieving volume-level Area Under ROC Curve (AUC) of 0.977 on HK dataset and 0.933 on Stanford dataset, respectively. The experimental results indicate the great potential of the proposed approach for the automated diagnosis system.
Subject(s)
Glaucoma , Tomography, Optical Coherence , Diagnostic Techniques, Ophthalmological , Glaucoma/diagnostic imaging , Humans , Supervised Machine Learning , Visual FieldsABSTRACT
BACKGROUND: Spectral-domain optical coherence tomography (SDOCT) can be used to detect glaucomatous optic neuropathy, but human expertise in interpretation of SDOCT is limited. We aimed to develop and validate a three-dimensional (3D) deep-learning system using SDOCT volumes to detect glaucomatous optic neuropathy. METHODS: We retrospectively collected a dataset including 4877 SDOCT volumes of optic disc cube for training (60%), testing (20%), and primary validation (20%) from electronic medical and research records at the Chinese University of Hong Kong Eye Centre (Hong Kong, China) and the Hong Kong Eye Hospital (Hong Kong, China). Residual network was used to build the 3D deep-learning system. Three independent datasets (two from Hong Kong and one from Stanford, CA, USA), including 546, 267, and 1231 SDOCT volumes, respectively, were used for external validation of the deep-learning system. Volumes were labelled as having or not having glaucomatous optic neuropathy according to the criteria of retinal nerve fibre layer thinning on reliable SDOCT images with position-correlated visual field defect. Heatmaps were generated for qualitative assessments. FINDINGS: 6921 SDOCT volumes from 1â384â200 two-dimensional cross-sectional scans were studied. The 3D deep-learning system had an area under the receiver operation characteristics curve (AUROC) of 0·969 (95% CI 0·960-0·976), sensitivity of 89% (95% CI 83-93), specificity of 96% (92-99), and accuracy of 91% (89-93) in the primary validation, outperforming a two-dimensional deep-learning system that was trained on en face fundus images (AUROC 0·921 [0·905-0·937]; p<0·0001). The 3D deep-learning system performed similarly in the external validation datasets, with AUROCs of 0·893-0·897, sensitivities of 78-90%, specificities of 79-86%, and accuracies of 80-86%. The heatmaps of glaucomatous optic neuropathy showed that the learned features by the 3D deep-learning system used for detection of glaucomatous optic neuropathy were similar to those used by clinicians. INTERPRETATION: The proposed 3D deep-learning system performed well in detection of glaucomatous optic neuropathy in both primary and external validations. Further prospective studies are needed to estimate the incremental cost-effectiveness of incorporation of an artificial intelligence-based model for glaucoma screening. FUNDING: Hong Kong Research Grants Council.