Optical coherence tomography angiography in the diagnosis of ocular disease

Clinical imaging provided by optical coherence tomography (OCT) and its variant, OCT-angiography (OCT-A), has revolutionised eyecare practice. The imaging techniques allow for the identification and quantification of ocular structures, supporting the diagnosis and prognosis of eye disease. In this review, an overview of the usefulness of OCT-A imaging in the diagnosis and management of a range of ocular conditions is provided when used in isolation or in combination with other imaging modalities and measures of visual function (visual field results). OCT-A imaging has the capacity to identify and quantify ocular vasculature non-invasively, thereby assisting the clinician in the diagnosis or to determine the efficacy of intervention in ocular conditions impacting retinal vasculature. Thus, additional clinically useful information can be obtained in eye diseases involving conditions such as those impacting retinal vessel occlusion, in diabetic retinopathy, inherited retinal dystrophy, age-related macular degeneration, choroidal neovascularisation and optic nerve disorders. Through a clinical case series, various ocular conditions are reviewed, and the impact of OCT-A imaging is discussed. Although OCT-A imaging has great promise and is already used in clinical management, there is a lack of set standards to characterise altered vascular features in disease and consequently for prognostication, primarily due to a lack of large-scale clinical trials and variability in OCT-A algorithms when generating quantitative parameters.

Two-Photon and Multiphoton Microscopy in Anterior Segment Diseases of the Eye

Two-photon excitation microscopy (TPM) and multiphoton fluorescence microscopy (MPM) are advanced forms of intravital high-resolution functional microscopy techniques that allow for the imaging of dynamic molecular processes and resolve features of the biological tissues of interest. Due to the cornea's optical properties and the uniquely accessible position of the globe, it is possible to image cells and tissues longitudinally to investigate ocular surface physiology and disease. MPM can also be used for the in vitro investigation of biological processes and drug kinetics in ocular tissues. In corneal immunology, performed via the use of TPM, cells thought to be intraepithelial dendritic cells are found to resemble tissue-resident memory T cells, and reporter mice with labeled plasmacytoid dendritic cells are imaged to understand the protective antiviral defenses of the eye. In mice with limbal progenitor cells labeled by reporters, the kinetics and localization of corneal epithelial replenishment are evaluated to advance stem cell biology. In studies of the conjunctiva and sclera, the use of such imaging together with second harmonic generation allows for the delineation of matrix wound healing, especially following glaucoma surgery. In conclusion, these imaging models play a pivotal role in the progress of ocular surface science and translational research.

Non-Invasive High-Resolution Imaging of In Vivo Human Myelinated Axons

This work aims to reveal the microscopic (2-3 micrometer resolution) appearance of human myelinated nerve fibers in vivo for the first time. We analyzed the myelinated retinal nerve fibers of a male patient without other neurological disorders in a non-invasive way using the transscleral optical phase imaging method with adaptive optics. We also analyzed the fellow eye with non-myelinated nerve fibers and compared the results with traditional ocular imaging methods such as optical coherence tomography. We documented the microscopic appearance of human myelin and myelinated axons in vivo. This method allowed us to obtain better details than through traditional ocular imaging methods. We hope these findings will be useful to the scientific community to evaluate neuro-retinal structures through new imaging techniques and more accurately document nerve anatomy and the pathophysiology of this disease.

Anterior-Segment Optical Coherence Tomography Unlocks Novel Perspectives: Lacking Iris Anterior Limiting Layer Signal in Uveitis

Chronic uveitis, a challenging intraocular inflammatory condition, presents complexities in diagnosis and management due to its diverse etiologies and manifestations. Anterior-segment optical coherence tomography (AS-OCT) has emerged as a pivotal tool in evaluating uveitis, offering high-resolution imaging of anterior segment structures. We present the case of a 49-year-old man diagnosed with neurosarcoidosis and chronic intermediate uveitis, where AS-OCT revealed unique findings. Clinical examination identified a Koeppe nodule. AS-OCT evaluation unveiled hyperreflectivity in the iris stroma and the adjacent nodule. Notably, AS-OCT documented the absence of the hyperreflective anterior limiting layer signal, a novel observation in uveitis assessment. This unprecedented finding underscores the significance of AS-OCT in elucidating uveitis pathophysiology and emphasizes its potential in refining diagnostic and therapeutic strategies for this complex ocular condition.

Progression of Rare Inherited Retinal Dystrophies May Be Monitored by Adaptive Optics Imaging

Inherited retinal dystrophies (IRDs) are bilateral genetic conditions of the retina, leading to irreversible vision loss. This study included 55 eyes afflicted with IRDs affecting the macula. The diseases examined encompassed Stargardt disease (STGD), cone dystrophy (CD), and cone-rod dystrophy (CRD) using adaptive optics (Rtx1™; Imagine Eyes, Orsay, France). Adaptive optics facilitate high-quality visualisation of retinal microstructures, including cones. Cone parameters, such as cone density (DM), cone spacing (SM), and regularity (REG), were analysed. The best corrected visual acuity (BCVA) was assessed as well. Examinations were performed twice over a 6-year observation period. A significant change was observed in DM (1282.73/mm2 vs. 10,073.42/mm2, p< 0.001) and SM (9.83 μm vs. 12.16 μm, p< 0.001) during the follow-up. BCVA deterioration was also significant (0.16 vs. 0.12, p = 0.001), albeit uncorrelated with the change in cone parameters. No significant difference in REG was detected between the initial examination and the follow-up (p = 0.089).

Segmentation-Free OCT-Volume-Based Deep Learning Model Improves Pointwise Visual Field Sensitivity Estimation

Purpose

The structural changes measured by optical coherence tomography (OCT) are related to functional changes in visual fields (VFs). This study aims to accurately assess the structure-function relationship and overcome the challenges brought by the minimal measurable level (floor effect) of segmentation-dependent OCT measurements commonly used in prior studies.

Methods

We developed a deep learning model to estimate the functional performance directly from three-dimensional (3D) OCT volumes and compared it to the model trained with segmentation-dependent two-dimensional (2D) OCT thickness maps. Moreover, we proposed a gradient loss to utilize the spatial information of VFs.

Results

Our 3D model was significantly better than the 2D model both globally and pointwise regarding both mean absolute error (MAE = 3.11 + 3.54 vs. 3.47 ± 3.75 dB, P < 0.001) and Pearson's correlation coefficient (0.80 vs. 0.75, P < 0.001). On a subset of test data with floor effects, the 3D model showed less influence from floor effects than the 2D model (MAE = 5.24 ± 3.99 vs. 6.34 ± 4.58 dB, P < 0.001, and correlation 0.83 vs. 0.74, P < 0.001). The gradient loss improved the estimation error for low-sensitivity values. Furthermore, our 3D model outperformed all prior studies.

Conclusions

By providing a better quantitative model to encapsulate the structure-function relationship more accurately, our method may help deriving VF test surrogates.

Translational Relevance

DL-based VF surrogates not only benefit patients by reducing the testing time of VFs but also allow clinicians to make clinical judgments without the inherent limitations of VFs.

Functional Carbon Quantum Dots for Ocular Imaging and Therapeutic Applications

Carbon quantum dots (CDs) are a class of emerging carbonaceous nanomaterials that have received considerable attention due to their excellent fluorescent properties, extremely small size, ability to penetrate cells and tissues, ease of synthesis, surface modification, low cytotoxicity, and superior water dispersion. In light of these properties, CDs are extensively investigated as candidates for bioimaging probes, efficient drug carriers, and disease diagnostics. Functionalized CDs represent a promising therapeutic candidate for ocular diseases. Here, this work reviews the potential use of functionalized CDs in the diagnosis and treatment of eye-related diseases, including the treatment of macular and anterior segment diseases, as well as targeting Aβ amyloids in the retina.

Federated Learning in Ocular Imaging: Current Progress and Future Direction

Advances in artificial intelligence deep learning (DL) have made tremendous impacts on the field of ocular imaging over the last few years. Specifically, DL has been utilised to detect and classify various ocular diseases on retinal photographs, optical coherence tomography (OCT) images, and OCT-angiography images. In order to achieve good robustness and generalisability of model performance, DL training strategies traditionally require extensive and diverse training datasets from various sites to be transferred and pooled into a "centralised location". However, such a data transferring process could raise practical concerns related to data security and patient privacy. Federated learning (FL) is a distributed collaborative learning paradigm which enables the coordination of multiple collaborators without the need for sharing confidential data. This distributed training approach has great potential to ensure data privacy among different institutions and reduce the potential risk of data leakage from data pooling or centralisation. This review article aims to introduce the concept of FL, provide current evidence of FL in ocular imaging, and discuss potential challenges as well as future applications.

Normative data of retinal arteriolar and venular calibre measurements determined using confocal scanning laser ophthalmoscopy system - Importance and implications for study of cardiometabolic disorders

Purpose:

To determine and validate retinal vascular caliber measurements by using the confocal scanning laser ophthalmoscopy system. Retinal vasculature changes are often regarded as clinical markers for systemic disease.

Methods:

It was a prospective observational study conducted on 600 eyes of 300 normal subjects with no systemic or ocular illness from January 1, 2016 to June 30, 2017 in a tertiary referral eye center. Non-mydriatic infrared reflectance, blue reflectance, and blue peak blue autofluorescence fundus imaging were done on the confocal scanning laser ophthalmoscopy system. The dimensions of the retinal vessels were measured using inbuilt calipers at 1800 mm from the center of the optic disc. Internal and external dimensions were measured. Observer variation and its comparison using Image J software were assessed.

Results:

The median age was 29 years (18–50 years). Mean internal and external diameters for arterioles were 85.1 ± 12.4 mm and 105.0 ± 12.0 mm, and for venules were 133.8 ± 16.6 mm and 145.4 ± 16.1 mm, respectively. The mean internal and external wall thicknesses were 19.7 ± 8.0 mm and 11.0 ± 5.6 mm, and wall thickness-to-lumen ratios were 0.3 ± 0.1 and 0.1 ± 0.1, respectively. Arteriolar-to-venular ratio for lumen and vessel was 0.66 ± 0.1 and 0.74 ± 0.1, respectively. There was no statistically significant difference between age groups. Both inter- and intra-observer reproducibility was >95%. The Bland–Altman plot showed that the difference between measurements using both confocal scanning laser ophthalmoscopy and Image J software lies within the limits of agreement approximately 95% of the time.

Conclusion:

This is the first effort to develop a normative database by using a simple non-invasive confocal scanning laser ophthalmoscopy system with high observer reproducibility.

In-Vivo Measurement of Ocular Deformation in Response to Ambient Pressure Modulation

A novel approach is presented for the non-invasive quantification of axial displacement and strain in corneal and anterior crystalline lens tissue in response to a homogenous ambient pressure change. A spectral domain optical coherence tomography (OCT) system was combined with a custom-built set of swimming goggles and a pressure control unit to acquire repetitive cross-sectional scans of the anterior ocular segment before, during and after ambient pressure modulation. The potential of the technique is demonstrated in vivo in a healthy human subject. The quantification of the dynamic deformation response, consisting of axial displacement and strain, demonstrated an initial retraction of the eye globe (−0.43 to −1.22 nm) and a subsequent forward motion (1.99 nm) in response to the pressure change, which went along with a compressive strain induced in the anterior crystalline lens (−0.009) and a tensile strain induced in the cornea (0.014). These mechanical responses appear to be the result of a combination of whole eye motion and eye globe expansion. The latter simulates a close-to-physiologic variation of the intraocular pressure and makes the detected mechanical responses potentially relevant for clinical follow-up and pre-surgical screening. The presented measurements are a proof-of-concept that non-contact low-amplitude ambient pressure modulation induces tissue displacement and strain that is detectable in vivo with OCT. To take full advantage of the high spatial resolution this imaging technique could offer, further software and hardware optimization will be necessary to overcome the current limitation of involuntary eye motions.

Imaging characteristics of the eyes of cinereous vulture (Aegypius monachus): morphology and comparative biometric measurement

The aim of this study is to describe radiographic, ultrasonographic, and computed tomographic appearance of normal cinereous vulture’s eye and to determine normal biometric values of intraocular structures. Twenty-six eyes of thirteen healthy cinereous vultures were examined. Under general anesthesia with isoflurane, ultrasonography (US), computed tomography (CT) and skull radiography were performed. Differences between both eyes as well as between US and CT measurements were investigated and correlation of measurements between both eyes as well as correlation between CT and US measurements of the various ocular structures were calculated. Most of paired data did not show any significant differences between both eyes and the CT and US measurements, while there were significant differences (P<0.05) between CT and US measurements of depth of both vitreous and anterior chambers, and axial length of the lens in right eyes. There was also a significant difference (P<0.05) between both eyes in depth of vitreous measured by CT. All the measurements had strong correlations between both eyes and between US and CT. In conclusion, ocular imaging techniques provided useful data of biometry and morphology, showing good correlation between CT and US in cinereous vulture’s eye. Especially, when ophthalmoscopic examinations would not be available due to opaque anterior segment, imaging techniques could be essential for diagnosing and managing of the eye.

High-Resolution Optical Coherence Tomography Angiography Characteristics of Limbal Stem Cell Deficiency

This study aimed to identify the anterior segment high-resolution optical coherence tomography (HR-OCT) and HR-OCT angiography (HR-OCTA) features suggestive of limbal stem cell deficiency (LSCD) as confirmed by both impression cytology (IC) and in vivo confocal microscopy (IVCM). This was a single-centre prospective cross-sectional study including 24 eyes of 22 patients with clinical suspicion of LSCD based on peripheral superficial corneal vascularisation and scarring. On IC and IVCM, performed and interpreted by blinded observers, 12 eyes each were diagnosed with and without LSCD. Additionally, 10 eyes of 5 healthy volunteers with no ocular pathology were also imaged. The 136 HR-OCT/A images of these 34 eyes were analysed with respect to 12 imaging parameters; the parameters most suggestive of LSCD were identified and the sensitivity and specificity were calculated. In the LSCD group, the most common aetiology was ocular chemical burns (83%), whereas in the non-LSCD group, the most common aetiology was viral keratitis (67%). Multiple logistic regression analysis revealed that mean epithelial reflectivity, mean stromal reflectivity, and mean superficial vascular density were the parameters that were diagnostic of LSCD on HR-OCT/A (p < 0.0001). A ratio of the mean epithelial reflectivity to stromal reflectivity of >1.29 corresponded with a high sensitivity (91.7%) and specificity (98.75%); while a mean superficial vascular density score of >0.38 corresponded with a sensitivity of 97.9% and specificity of 73.8%. In conclusion, HR-OCT/A as a non-invasive imaging modality could prove to be a useful tool for confirming the diagnosis of LSCD, with potential clinical and research applications.

Association of Cognitive Function and Retinal Neural and Vascular Structure in Type 1 Diabetes

CONTEXT

Cognitive dysfunction is a growing and understudied public health issue in the aging type 1 diabetic population and is difficult and time-consuming to diagnose. Studies in long duration type 1 diabetes have reported the presence of proliferative diabetic retinopathy was associated with cognitive dysfunction.

OBJECTIVE

This study assessed whether structural and vascular abnormalities of the retina, representing an extension of the central nervous system, are associated with cognitive impairment and other complications of type 1 diabetes.

METHODS

An observational cross-sectional study of individuals with 50 or more years of type 1 diabetes (Joslin Medalist Study) was conducted at a university hospital in the United States. The study included 129 participants with complete cognitive testing. Validated cognitive testing measures included psychomotor speed, and immediate, and delayed memory. Optical coherence tomography (OCT) and OCT angiography (OCTA) were performed to obtain neural retinal layer thicknesses and vascular density for superficial (SCP) and deep retinal capillary plexus (DCP). Multivariable modeling was adjusted for potential confounders associated with outcomes in unadjusted analyses.

RESULTS

Decreased vessel density of the SCP and DCP was associated with worse delayed memory (DCP: P = .002) and dominant hand psychomotor speed (SCP: P = .01). Thinning of the retinal outer nuclear layer was associated with worse psychomotor speed both in nondominant and dominant hands (P = .01 and P = .05, respectively). Outer plexiform layer thickness was associated with delayed memory (P = .04).

CONCLUSION

These findings suggest that noninvasive retinal imaging using OCT and OCTA may assist in estimating the risks for cognitive dysfunction in people with type 1 diabetes.

Deep Learning for Anterior Segment Optical Coherence Tomography to Predict the Presence of Plateau Iris

Purpose

The purpose of this study was to evaluate the diagnostic performance of deep learning (DL) anterior segment optical coherence tomography (AS-OCT) as a plateau iris prediction model.

Design

We used a cross-sectional study of the development and validation of the DL system.

Methods

We conducted a collaboration between a referral eye center and an informative technology department. The study enrolled 179 eyes from 142 patients with primary angle closure disease (PACD). All patients had remaining appositional angle after iridotomy. Each eye was scanned in four quadrants for both AS-OCT and ultrasound biomicroscopy (UBM). A DL algorithm for plateau iris prediction of AS-OCT was developed from training datasets and was validated in test sets. Sensitivity, specificity, and area under the receiver operating characteristics curve (AUC-ROC) of the DL for predicting plateau iris were evaluated, using UBM as a reference standard.

Results

Total paired images of AS-OCT and UBM were from 716 quadrants. Plateau iris was observed with UBM in 276 (38.5%) quadrants. Trainings dataset with data augmentation were used to develop an algorithm from 2500 images, and the test set was validated from 160 images. AUC-ROC was 0.95 (95% confidence interval [CI] = 0.91 to 0.99), sensitivity was 87.9%, and specificity was 97.6%.

Conclusions

DL revealed a high performance in predicting plateau iris on the noncontact AS-OCT images.

Translational Relevance

This work could potentially assist clinicians in more practically detecting this nonpupillary block mechanism of PACD.

Carbon Dots Fabrication: Ocular Imaging and Therapeutic Potential

Vision loss is a major complication in common ocular infections and diseases such as bacterial keratitis, age-related macular degeneration (AMD) and diabetic retinopathy (DR). The prevalence of such ophthalmic diseases represents an urgent need to develop safe, effective, and long-term treatments. Current therapies are riddled with drawbacks and limitations which calls for the exploration of alternative drug delivery mechanisms. Toxicity of the inorganic metals and metal oxides used for drug delivery raise safety concerns that are alleviated with the alternate use of, a natural and organic polymer which is both biocompatible and environmentally friendly. Carbon dots (CDs) represent a great potential in novel biomedical applications due to their tunable fluorescence, biocompatibility, and ability to be conjugated with diverse therapeutic materials. There is a growing interest on the exploitation of these properties for drug delivery with enhanced bio-imaging. However, there are limited reports of CD applications for ophthalmic indications. In this review, we focus on the CD potential and the development of translational therapies for ophthalmic diseases. The current review presents better understanding of fabrication of CDs and how it may be useful in delivering anti-bacterial agents, anti-VEGF molecules as well as imaging for ophthalmic applications.

Development of Deep Learning Models to Predict Best-Corrected Visual Acuity from Optical Coherence Tomography

Purpose

To develop deep learning (DL) models to predict best-corrected visual acuity (BCVA) from optical coherence tomography (OCT) images from patients with neovascular age-related macular degeneration (nAMD).

Methods

Retrospective analysis of OCT images and associated BCVA measurements from the phase 3 HARBOR trial (NCT00891735). DL regression models were developed to predict BCVA at the concurrent visit and 12 months from baseline using OCT images. Binary classification models were developed to predict BCVA of Snellen equivalent of <20/40, <20/60, and ≤20/200 at the concurrent visit and 12 months from baseline.

Results

The regression model to predict BCVA at the concurrent visit had R² = 0.67 (root-mean-square error [RMSE] = 8.60) in study eyes and R² = 0.84 (RMSE = 9.01) in fellow eyes. The best classification model to predict BCVA at the concurrent visit had an area under the receiver operating characteristic curve (AUC) of 0.92 in study eyes and 0.98 in fellow eyes. The regression model to predict BCVA at month 12 using baseline OCT had R² = 0.33 (RMSE = 14.16) in study eyes and R² = 0.75 (RMSE = 11.27) in fellow eyes. The best classification model to predict BCVA at month 12 had AUC = 0.84 in study eyes and AUC = 0.96 in fellow eyes.

Conclusions

DL shows promise in predicting BCVA from OCTs in nAMD. Further research should elucidate the utility of models in clinical settings.

Translational Relevance

DL models predicting BCVA could be used to enhance understanding of structure–function relationships and develop more efficient clinical trials.

Approaches to quantify optical coherence tomography angiography metrics

Optical coherence tomography (OCT) has revolutionized the field of ophthalmology in the last three decades. As an OCT extension, OCT angiography (OCTA) utilizes a fast OCT system to detect motion contrast in ocular tissue and provides a three-dimensional representation of the ocular vasculature in a non-invasive, dye-free manner. The first OCT machine equipped with OCTA function was approved by U.S. Food and Drug Administration in 2016 and now it is widely applied in clinics. To date, numerous methods have been developed to aid OCTA interpretation and quantification. In this review, we focused on the workflow of OCTA-based interpretation, beginning from the generation of the OCTA images using signal decorrelation, which we divided into intensity-based, phase-based and phasor-based methods. We further discussed methods used to address image artifacts that are commonly observed in clinical settings, to the algorithms for image enhancement, binarization, and OCTA metrics extraction. We believe a better grasp of these technical aspects of OCTA will enhance the understanding of the technology and its potential application in disease diagnosis and management. Moreover, future studies will also explore the use of ocular OCTA as a window to link ocular vasculature to the function of other organs such as the kidney and brain.

Ultralow energy photoacoustic microscopy for ocular imaging in vivo

SIGNIFICANCE

The development of ultralow energy photoacoustic microscopy (PAM) on the clinically relevant pigmented rabbit eye model paves a road toward translation of the emerging PAM technology in ophthalmology clinics.

AIM

Since the eye is particularly vulnerable to laser damage, we aim to develop an ultralow energy PAM system to significantly improve the laser safety of PAM by increasing the sensitivity of the system and reducing the incident laser energy for imaging.

APPROACH

A multichannel data acquisition circuit with two-stage signal amplification was specially designed, which, in combination with the application of 3 by 3 median filter in the spatial domain, significantly improved the signal-to-noise ratio of the PAM system. The safety of this system was validated by histopathology, fluorescein angiography, and fundus photography.

RESULTS

Experiments performed on pigmented rabbits demonstrated that, when using this ultralow energy PAM system, satisfactory image quality can be achieved in the eye with an incident laser fluence that is only 1% of the American National Standards Institute safety limit. Fundus photography, fluorescein angiography, and histopathology were performed after the imaging procedure, and no retinal or ocular damage was observed.

CONCLUSIONS

The proposed ultralow energy PAM system has excellent safety and holds potential to be developed into a clinical tool for ocular imaging.

Noninvasive and Noncontact Sequential Imaging of the Iridocorneal Angle and the Cornea of the Eye

Purpose

High-resolution imaging of the critical anatomic structures of the eye, especially of the anterior chamber, in vivo, remains a challenge, even with currently available state-of-the-art medical imaging techniques. This study aims for the noninvasive and noncontact sequential imaging of the iridocorneal angle, especially the trabecular meshwork (TM) and the cornea of the eye in high-resolution using a newly developed imaging platform.

Methods

Bessel beam scanned light sheet fluorescence microscopy is used to attain high-resolution images of the TM. The ability of the Bessel beam to self-reconstruct around obstacles increases the image contrast at the TM region inside eye by reducing scattering and shadow artifacts. With minimal modifications, the excitation arm of the developed imaging system is adapted for noncontact, high-resolution corneal imaging.

Results

High-resolution images of the TM structures and cellular-level corneal structures are obtained in ex vivo porcine eyes, and subsequently in New Zealand white rabbit, in vivo. The spatial resolution of the developed system is 2.19 µm and has a noncontact working distance of 20 mm.

Conclusions

A high-resolution imaging platform for noncontact sequential imaging of the TM and the cornea of the eye is developed. This imaging system is expected to be of potential interest in the evaluation and diagnosis of glaucoma and corneal diseases.

Translational Relevance

The developed prototype offers the plausibility of in vivo, noncontact, and high-resolution imaging of the iridocorneal angle and cornea of the eye that will aid clinicians in diagnosing open-angle glaucoma and corneal diseases better.

Health Economic and Safety Considerations for Artificial Intelligence Applications in Diabetic Retinopathy Screening

Systematic screening for diabetic retinopathy (DR) has been widely recommended for early detection in patients with diabetes to address preventable vision loss. However, substantial manpower and financial resources are required to deploy opportunistic screening and transition to systematic DR screening programs. The advent of artificial intelligence (AI) technologies may improve access and reduce the financial burden for DR screening while maintaining comparable or enhanced clinical effectiveness. To deploy an AI-based DR screening program in a real-world setting, it is imperative that health economic assessment (HEA) and patient safety analyses are conducted to guide appropriate allocation of resources and design safe, reliable systems. Few studies published to date include these considerations when integrating AI-based solutions into DR screening programs. In this article, we provide an overview of the current state-of-the-art of AI technology (focusing on deep learning systems), followed by an appraisal of existing literature on the applications of AI in ophthalmology. We also discuss practical considerations that drive the development of a successful DR screening program, such as the implications of false-positive or false-negative results and image gradeability. Finally, we examine different plausible methods for HEA and safety analyses that can be used to assess concerns regarding AI-based screening.

Optical sectioning and high resolution visualization of trabecular meshwork using Bessel beam assisted light sheet fluorescence microscopy

Glaucoma, one of the leading causes of blindness, is an eye disease caused by irregularities in the ocular aqueous outflow system causing an elevated intraocular pressure. High resolution imaging of the aqueous outflow system comprising trabecular meshwork is immensely valuable to vision analysts and clinicians in comprehending the disease state for the efficacious analysis and treatment of glaucoma. Currently available ocular imaging devices are unable to deliver high resolution images for the visualization of the trabecular meshwork. A method to obtain high resolution (sub-micrometer) images of the trabecular meshwork using Bessel-Gauss beam scanned light sheet fluorescence microscopy is presented and the optical sectioning capability of this technique to obtain three-dimensional volumetric images of the trabecular meshwork of an intact eye without any physical dissection is demonstrated. Figure: Three-dimensional visualization of trabecular meshwork of porcine eye.

Quantitative Evaluation of Gonioscopic and EyeCam Assessments of Angle Dimensions Using Anterior Segment Optical Coherence Tomography

Purpose

To evaluate the relationship between angle dimensions assessed by gonioscopy or EyeCam and anterior segment optical coherence tomography (AS-OCT).

Methods

Subjects aged 50 years or older were recruited from the Chinese American Eye Study (CHES). Each subject underwent a complete ocular exam, including gonioscopy, AS-OCT, and EyeCam. Angle closure was defined as three or more quadrants in which pigmented trabecular meshwork could not be visualized. Angle opening distance (AOD), angle recess area (ARA), trabecular iris space area (TISA), trabecular iris angle (TIA), and scleral spur angle (SSA) were measured in each AS-OCT image.

Results

709 eyes (272 angle closure, 437 open angle) from 709 subjects were analyzed. Mean gonioscopy and EyeCam grades tended to increase as AS-OCT measurements increased. There were strong correlations overall between AS-OCT measurements and gonioscopy (r > 0.73) and EyeCam (r > 0.68) grades. However, correlations with AS-OCT measurements were weak for gonioscopy (r < 0.38) and EyeCam (r < to 0.27) among eyes with angle closure. Mean AS-OCT measurements differed for eyes with Shaffer grade 0 in all four quadrants among eyes with varying degrees of angle closure on gonioscopy (P < 0.01) but did not differ among eyes with varying degrees of angle closure on EyeCam (P > 0.27).

Conclusions

Angle assessments by gonioscopy and EyeCam are weakly related to angle dimensions in eyes with angle closure.

Translational Relevance

AS-OCT imaging raises concerns about current clinical methods that rely on direct visualization of ACA structures to assess the degree of angle closure.

Noninvasive Assessment of Retinal Blood Flow Using a Novel Handheld Laser Speckle Contrast Imager

Purpose

We assessed the image quality and reproducibility of blood flow measurements from a novel handheld laser speckle imager in handheld and stabilized use cases.

Methods

Eleven dilated human subjects were imaged with the XyCAM Handheld Retinal Imager investigational device (XyCAM HRI) in the handheld and stabilized use case in nine consecutive imaging sessions. Subjects then underwent standard color fundus photography using a Topcon TRC 50DX. The vessel-to-background contrast of the XyCAM HRI red-free photo was compared to the fundus photograph, while the coefficient of variation of blood flow measurements in specific arteries and veins also was determined.

Results

Vessel-to-background contrast was statistically greater in the handheld use case when compared to the standard color fundus photographs (P = 0.01). Estimates of mean blood flow velocity (BFV) were highly correlated between the stabilized and handheld use case (r² = 0.96). Peak velocity estimates in arteries were significantly higher than those in veins (P < 0.05).

Conclusions

The XyCAM HRI prototype can acquire fundus photographs with the same or better level of clarity as color fundus photographs, and reproducibly acquire functional blood flow information in the handheld use case.

Translational Relevance

To our knowledge, this is the first human study of a handheld laser speckle retinal imaging device. Determination of retinal blood flow has applications to ophthalmic and systemic disease and a portable handheld retinal imager that determines blood flow may be widely adopted at the point of care.

Cartesian MR fingerprinting in the eye at 7T using compressed sensing and matrix completion-based reconstructions

PURPOSE

To explore the feasibility of MR Fingerprinting (MRF) to rapidly quantify relaxation times in the human eye at 7T, and to provide a data acquisition and processing framework for future tissue characterization in eye tumor patients.

METHODS

In this single-element receive coil MRF approach with Cartesian sampling, undersampling is used to shorten scan time and, therefore, to reduce the degree of motion artifacts. For reconstruction, approaches based on compressed sensing (CS) and matrix completion (MC) were used, while their effects on the quality of the MRF parameter maps were studied in simulations and experiments. Average relaxation times in the eye were measured in 6 healthy volunteers. One uveal melanoma patient was included to show the feasibility of MRF in a clinical context.

RESULTS

Simulation results showed that an MC-based reconstruction enables large undersampling factors and also results in more accurate parameter maps compared with using CS. Experiments in 6 healthy volunteers used a reduction in scan time from 7:02 to 1:16 min, producing images without visible loss of detail in the parameter maps when using the MC-based reconstruction. Relaxation times from 6 healthy volunteers are in agreement with values obtained from fully sampled scans and values in literature, and parameter maps in a uveal melanoma patient show clear difference in relaxation times between tumor and healthy tissue.

CONCLUSION

Cartesian-based MRF is feasible in the eye at 7T. High undersampling factors can be achieved by means of MC, significantly shortening scan time and increasing patient comfort, while also mitigating the risk of motion artifacts.

Measurement of the hypotenuse of the vertical optic nerve head cup with spectral-domain optical coherence tomography for the structural diagnosis of glaucoma

Purpose

To evaluate the hypotenuse of the vertical optic nerve head cup (HVOC), measured using the length and depth of the cup obtained with enhanced depth imaging spectral-domain optic coherence tomography (SD-OCT), as a biomarker for glaucoma diagnosis.

Methods

This was a prospective cross-sectional study of patients with glaucoma and controls. SD-OCT was performed in all participants to assess average circumpapillary retinal nerve fiber layer (RNFL) thickness. A vertical B-scan of the optic nerve head (ONH) was obtained for HVOC measurement. The length and depth of the optic nerve cup formed the sides of a right triangle that were used to calculate the HVOC. Participants also underwent standard automated perimetry.

Results

One hundred and fifty-six eyes were divided into three groups: mean deviation (MD) <−7 dB (60 eyes); MD ≥−7 dB (74 eyes); and healthy subjects (22 eyes). The mean (SD) HVOC in these groups was 1,419.8 (347.2) µm, 1,234.6 (258.8) µm, and 685.79 (315.4) µm (P<0.01), respectively. In the secondary structure–function analysis, only discs with a vertical diameter of 1.51–2.00 mm were included (120 eyes). The HVOCs were divided into four percentile groups, with the following means: 940, 1,128, 1,390, and 1,662 µm. There was a significant difference in MD between percentile groups 1 and 3 (P<0.03), 1 and 4 (P<0.001), 2 and 3 (P<0.02), and 2 and 4 (P<0.001). RNFL thickness differed among all percentile groups (P<0.001).

Conclusion

HVOC may provide an additional morphometric biomarker for the structural evaluation of ONH remodeling in glaucoma.

Factors affecting signal strength in spectral-domain optical coherence tomography

PURPOSE

To identify ocular factors that affect signal strength in spectral-domain optical coherence tomography (SD-OCT).

METHODS

Data from 1312 participants of the population-based Singapore Malay Eye Study-2 (SiMES-2) were included in the analysis. All participants underwent standardized ophthalmic examination, including measurements of best-corrected visual acuity (BCVA), refractive error, axial length, corneal curvature and presence of cataracts. Optic disc and macular cube scans were acquired using the Cirrus HD-OCT (software version 6.0, Carl Zeiss Meditec, Dublin, CA, USA). Signal strength of the optical coherence tomography (OCT) scan was recorded for each study eye. Multivariable linear regression analyses were performed to evaluate the associations between ocular factors and signal strength of the OCT scans.

RESULTS

The mean (±SD) age of our study participants was 61 ± 9 years, and 44.6% were male. Mean optic disc scan signal strength was 7.90 ± 1.25, range = 0-10, while mean macular scan signal strength was 8.80 ± 1.27, range = 0-10. In multivariable regression analyses, poorer signal strength in optic disc and macular cube scans was each associated with older age (per decade, β = -0.373, p < 0.001; β = -0.373, p < 0.001, respectively), poorer BCVA (per logMAR line; β = -0.123, p < 0.001; β = -0.156, p < 0.001, respectively), greater degree of myopia (per negative dioptre of spherical equivalent; β = -0.112, p < 0.001; β = -0.117, p < 0.001, respectively), presence of cortical cataracts (β = -0.331, p < 0.001; β = -0.314, p < 0.001, respectively) and presence of posterior subcapsular cataracts (β = -0.910, p < 0.001; β = -0.797, p < 0.001, respectively).

CONCLUSION

We found that older age, poorer BCVA, greater degree of myopia and presence of cortical and posterior subcapsular cataracts were associated with reduced signal strength in Cirrus SD-OCT. Our findings provide information on the barriers to obtaining good image quality when using SD-OCT, and allow clinicians to potentially identify individuals who are more likely to have unreliable OCT measurements.

Whole-body positional manipulators for ocular imaging of anaesthetised mice and rats: a do-it-yourself guide

Background

In vivo retinal imaging of rodents has gained a growing interest in ophthalmology and neurology. The bedding of the animals with the possibility to perform adjustments in order to obtain an ideal camera-to-eye angle is challenging.

Methods

We provide a guide for a cost-effective, do-it-yourself rodent holder for ocular imaging techniques. The set-up was tested and refined in over 2000 optical coherence tomography measurements of mice and rats.

Results

The recommended material is very affordable, readily available and easily assembled. The holder can be adapted to both mice and rats. A custom-made mouthpiece is provided for the use of inhalant anaesthesia. The holder is highly functional and assures that the rodent’s eye is the centre of rotation for adjustments in both the axial and the transverse planes with a major time benefit over unrestrained positioning of the rodents.

Conclusion

We believe this guide is very useful for eye researchers focusing on in vivo retinal imaging in rodents as it significantly reduces examination times for ocular imaging.

Hybrid-modality ocular imaging using a clinical ultrasound system and nanosecond pulsed laser

Hybrid optical modality imaging is a special type of multimodality imaging significantly used in the recent past in order to harness the strengths of different imaging methods as well as to furnish complementary information beyond that provided by any individual method. We present a hybrid-modality imaging system based on a commercial clinical ultrasound imaging (USI) system using a linear array ultrasound transducer (UST) and a tunable nanosecond pulsed laser as the source. The integrated system uses photoacoustic imaging (PAI) and USI for ocular imaging to provide the complementary absorption and structural information of the eye. In this system, B-mode images from PAI and USI are acquired at 10 Hz and about 40 Hz, respectively. A linear array UST makes the system much faster compared to other ocular imaging systems using a single-element UST to form B-mode images. The results show that the proposed instrumentation is able to incorporate PAI and USI in a single setup. The feasibility and efficiency of this developed probe system was illustrated by using enucleated pig eyes as test samples. It was demonstrated that PAI could successfully capture photoacoustic signals from the iris, anterior lens surface, and posterior pole, while USI could accomplish the mapping of the eye to reveal the structures like the cornea, anterior chamber, lens, iris, and posterior pole. This system and the proposed methodology are expected to enable ocular disease diagnostic applications and can be used as a preclinical imaging system.

Magnetic resonance imaging of the retina: from mice to men

This mini-review provides an overview of magnetic resonance imaging (MRI) applications to study rodent, cat, non-human primate, and human retinas. These techniques include T(1) - and T(2) -weighted anatomical, diffusion, blood flow, blood volume, blood-oxygenation level dependent, manganese-enhanced, physiological, and functional MRI. Applications to study the retinas in diabetic retinopathy, glaucoma, and retinal degeneration are also reviewed. MRI offers some unique advantages compared with existing imaging techniques and has the potential to further our understanding of physiology and function in healthy and diseased retinas.

Ultra-widefield fluorescein angiography of white without pressure

Purpose

To describe ultra-widefield fluorescein angiography (UWFA) findings in eyes with white without pressure (WWOP) and in eyes without any obvious peripheral chorioretinal disease, and to determine if a difference exists between these two groups.

Methods

A retrospective review of 379 eyes undergoing diagnostic UWFA using the Optos 200Tx imaging system. Eyes were excluded if the quality of the color photograph or UWFA prevented reliable evaluation. Eyes were also excluded if there was any evidence of peripheral retinal or choroidal disease, which was thought to have an effect on UWFA (eg, peripheral background diabetic or hypertensive retinopathy, vein occlusion, or any other peripheral vascular disorder). Eyes were determined to have WWOP, based on a dilated fundus examination and color fundus photography that contained areas of peripheral retinal whitening consistent with the diagnosis. UWFA was evaluated by trained masked graders, and determined to have or not have peripheral vascular leakage and/or staining.

Results

Of the 379 eyes evaluated, 45 eyes were included in the study. Twelve eyes were determined to have peripheral WWOP; 33 eyes did not have WWOP on examination or color fundus photography. Three common UWFA peripheral patterns were visualized. Eyes with and without WWOP were grouped into one of three patterns. The majority of eyes without WWOP demonstrated UWFA pattern one (69.7%), while those in the WWOP group demonstrated pattern three (50%). The distribution of UWFA patterns is statistically different between those with and without WWOP (P = 0.002). In eyes without WWOP, in patients with no documented systemic microvascular disease (diabetes, hypertension), 71.4% of eyes had UWFA pattern one while 14.3% had both patterns two and three.

Conclusion

This study is one of the first to specifically evaluate peripheral vascular leakage/staining in eyes with WWOP as well as in eyes without any obvious peripheral chorioretinal disease. We demonstrate that a significant portion of WWOP eyes exhibit peripheral findings on UWFA (pattern one) compared to eyes without WWOP. Importantly, even in eyes that are apparently unremarkable in the periphery on exam and color photography, UWFA can still show peripheral vascular abnormalities. These results warrant further investigation.

Correlation of a scanning laser derived oedema index and visual function following grid laser treatment for diabetic macular oedema

AIM

To correlate change of an oedema index derived by scanning laser tomography with change of visual function in patients undergoing grid laser photocoagulation for clinically significant diabetic macular oedema (DMO).

METHODS

The sample comprised 24 diabetic patients with retinal thickening within 500 micro m of the fovea. Inclusion criteria included a logMAR visual acuity of 0.25, or better. Patients were assessed twice before a single session of grid laser treatment and within 1 week of, and at 1, 2, 4, and 12 weeks after, treatment. At each visit, patients underwent logMAR visual acuity, conventional and short wavelength automated perimetry (SWAP), and scanning laser tomography. Each visual function parameter was correlated with the mean oedema index. The mean oedema index represented the z-profile signal width divided by the maximum reflectance intensity (arbitrary units). A Pearson correlation coefficient (Bonferroni corrected) was undertaken on the data set of each patient.

RESULTS

13 patients exhibited significant correlation of the mean oedema index and at least one measure of visual function for the 10 degrees x 10 degrees scan field while 10 patients correlated for the 20 degrees x 20 degrees scan field. Seven patients demonstrated correlation for both scan fields. Laser photocoagulation typically resulted in an immediate loss of perimetric sensitivity whereas the oedema index changed over a period of weeks. Localised oedema did not impact upon visual acuity or letter contrast sensitivity when situated extrafoveally.

CONCLUSIONS

Correlation of change of the oedema index and of visual function following grid laser photocoagulation was not found in all patients. An absence of correlation can be explained by the localised distribution of DMO in this sample of patients, as well as by differences in the time course of change of the oedema index and visual function. The study has objectively documented change in the magnitude and distribution of DMO following grid laser treatment and has established the relation of this change to the change in visual function.