This PDF file contains the front matter associated with SPIE Proceedings Volume 11623, including the Title Page, Copyright information, and Table of Contents.

Welcome and Introduction to SPIE Conference 11623: Ophthalmic Technologies XXXI.

Networking Session for Conference 11623: Ophthalmic Technologies XXXI

Introduction to Pascal Rol Lecture

Not going it alone: Physicians innovating with scientists and engineers

The highest three-dimensional (3D) resolution possible in in-vivo retinal imaging is achieved by combining optical coherence tomography (OCT) and adaptive optics. However, this combination brings important limitations, such as small field-of-view and complex, cumbersome systems, preventing so far the translation of this technology from the research lab to clinics. Here, we mitigate these limitations by combining our compact time-domain full-field OCT (FFOCT) with a multi-actuator adaptive lens positioned just in front of the eye, in a technique we call the adaptive-glasses wavefront sensorless approach. Through this approach, we demonstrate that ocular aberrations can be corrected, increasing the FFOCT signal-to-noise ratio and enabling imaging of different retinal layers with a 2μm x 2μm x 8μm resolution over a 5° x 5° field-of-view, without major anisoplanatism influence.

Adaptive optics (AO) enables retinal imaging at cellular resolution. Today, most ophthalmic AO systems have closed-loop bandwidths of ≤2 Hz, insufficient for many conditions encountered in the clinic. Here, we develop an ultrafast AO with a bandwidth of 32.6 Hz and evaluate its use with optical coherence tomography. After AO activation, the RMS wavefront aberration from an un-cyclopleged human eye dropped below diffraction limit within 5 ms, 40× faster than the fastest ophthalmic AO system reported in the literature. Because the system converges so quickly, we can use the data immediately after a blink or when imaging locations are changed, even in eyes wearing contact lenses.

We present Optical Incoherence Tomography (OIT): a completely digital method to generate tomographic retinal cross-sections from en-face through-focus image stacks acquired by non-interferometric imaging systems, such as en-face adaptive optics (AO)-ophthalmoscopes. We show how to use OIT to guide focus position in cases where the user is “blind" focusing, such as auto fluorescence imaging of the Retinal Pigment Epithelium (RPE). We also demonstrate that OIT can produce distinctive cross-sectional views of the retina using back-scattered, multiply-scattered or even fluorescent light, making it a complementary technique to OCT.

We introduce an anamorphic detection paradigm to optimize spatial and spectral resolution in adaptive optics line-scan OCT, wherein an improved light collection efficiency and signal roll-off compared to traditional methods was demonstrated. The benefits for in vivo imaging were exemplified by retrieving nanometer-scale light-induced optical path length dynamics at high speed in individual cones. The high speed, sensitivity and cellular-scale resolution of the resulting adaptive optics line-scan OCT instrument offers a robust and sensitive biomarker for retinal function in health and disease.

Adaptive optics optical coherence tomography technology enables non-invasive high-resolution retinal imaging and promises earlier detection of ocular disease. However, the images are corrupted by eye-movement artifacts that must be corrected to permit proper image analysis. We developed a method for efficiently removing eye-movement artifacts of A-lines using a multiple-reference global coordinate system. It corrects 3D translational eye movements, torsional eye movements, and image scaling, minimizing image distortion and substantially improving both regularity of the cone photoreceptor mosaic and clarity of individual cones.

High-resolution imaging is essential for understanding retinal diseases. Adaptive optics scanning light ophthalmoscopy (AOSLO) achieves cellular-level resolution through correction of the optical aberrations of the eye. However, the resolution of AOSLO is still limited by the diffraction of light. Here, we combine annular pupil illumination with sub-Airy disk confocal pinhole detection to surpass the diffraction limit. With the improved resolution, both rod photoreceptor and foveal cone mosaics were more readily identifiable in the living human eye.

An adaptive optics optical coherence tomography (AO-OCT) system with a non-modulated pyramid wavefront sensor (P-WFS) is presented. The P-WFS is implemented as add-on to a Shack Hartmann (SH) WFS based AO-OCT system, where the AO loop is driven by either sensor. Equivalent performance of the AO-OCT is demonstrated with the P-WFS and the SH-WFS by visualizing the mosaics of different retinal cell types in image data obtained in-vivo at the fovea of a healthy volunteer. Contrary to the SH-WFS, the pupil sampling of P-WFS is flexible and may be adjusted in order to further increase the sensitivity of the sensor.

In the ophthalmic Shack-Hartmann wavefront sensor beacon light originates from various depths within the retina. We model the retina of the human and mouse eyes as formed by two backscattering planes, showing that the use of lenslets with low Fresnel number and/or too small search boxes in the centroiding algorithm will produce artifactual aberrations. We evaluate the impact of these errors for four common beacon illumination strategies: full circular, annular, small circular on-axis and small circular off-axis. We find that artifactual aberrations are larger for annular and off-axis beacon illumination, dominated by defocus plus spherical aberration and defocus plus coma, respectively. These artifactual aberrations can be almost completely eliminated by selecting the minimum centroid search box size based on a simple Gaussian optics model and ocular biometry, provided the lenslet Fresnel number is sufficiently large to introduce minimal cross-talk between images of adjacent lenslets.

A new detection scheme was developed for simultaneous multi-channel imaging that provides isotropic images of retinal structures, free of directionality artifacts. The arrangement consists of light collecting fibers that act as offset apertures. This fiber bundle configuration can be used to retrofit basically any existing AO-SLO platform. The channels can be combined to reveal additional structural and functional details and this kind of retinal imaging with cellular resolution is a valuable new tool for researchers and clinicians.

Optical coherence tomography (OCT) is a well-established modality providing cross-sectional images of the human retina. OCT, however, does not provide high-resolution en face images of the outer retinal layers due to eye aberrations and the tradeoff between imaging depth and transverse resolution. Recently, we demonstrated the spatiotemporal optical coherence (STOC) that provides high-resolution, aberration-free volumetric images of the retina and cornea with the voxel rate of 8 GHz. Here, we show that STOC can be used for angiography, and allows us to estimate Doppler broadening, enabling visualization of the choroidal blood vessels in the human retina in vivo.

We developed a novel ophthalmic imaging platform that combines non-invasive measurements of retina/choroid structure and ocular blood flow based on optical coherence tomography (OCT) and wide-field semi-quantitative global flow visualization using line-scanning Doppler flowmetry (LSDF). The combination of these two imaging modalities within the same imaging platform enables comprehensive assessment of blood flow in retina and choroid and provides efficient characterization of blood flow in hemodynamic studies both in human volunteers and in small animals. The platform enables visualization of the entire posterior hemisphere vasculature, including vortex veins, using only light and without additional contrast agent in humans and rats.

We present a digital method to reveal the axial direction of local blood flow based on the Doppler spectrum asymmetry, called directional contrast. This contrast is overlaid on standard grayscale blood flow images to depict flow towards the camera in red, and flow away from the camera in blue. The local direction of blood flow with respect to the optical axis can be revealed with a high temporal resolution in out-of-plane retinal vessels, allowing to evidence potential blood flow reversal. We demonstrate this capability in an eye affected by high tension glaucoma and central retinal vein occlusion, in which we found a significant diastolic arterial retrograde flow.

Transient stoppage of erythrocytes through different vascular beds has important implications for local tissue metabolism. By combining adaptive optics retinal imaging with erythrocyte-mediated angiography (AO-EMA), erythrocyte stasis events can be readily observed in the microvasculature of living human eyes. Localization of erythrocyte stasis using EMA alongside AO-based indocyanine green (ICG) angiography illustrate the notion that there is a previously uncharacterized population of erythrocytes in stasis residing in the smallest choroidal vessels. These observations are an important step towards elucidating the hemodynamic properties of the choroidal microcirculation and demonstrate a novel application of ICG imaging.

Optical coherence tomography (OCT) sampling density and acquisition time is directly related to scanning mirror performance. A galvanometer impulse response that is broad or rings reduces the linearly-sampled regions of OCT B-scans and can increase total acquisition time, cause nonlinear distortions, and reduce resolution and contrast in OCT angiography (OCTA). Here, we utilize hardware and software optimizations to improve galvanometer frequency response in order to minimize scanner settling time and are thus able to improve imaging speed and maximize sampling linearity and field-of-view (FOV). We believe these methods can benefit both real-time retinal tracking in OCT and OCTA acquisition protocols.

Glaucoma is an optic neuropathy characterized by loss of retinal ganglion cells and their axons. Glaucoma has a strong vascular component and decreased macular vessel density is known to be associated with glaucomatous damage. Adaptive optics – optical coherence tomography allows for the simultaneous quantification of vascular and ganglion cell densities. We observed a moderately strong correlation between ganglion cell and vessel densities across the macula, as well as some correlation at individual locations. Vascular density may prove to be a useful surrogate biomarker of glaucoma progression and with further study reveal new information on impairment of neurovascular coupling in glaucoma.

Here we describe a new processing pipeline for assessing large amounts of OCT data and apply it to longitudinal imaging data from control and Superoxide Dismutase 1 knockout mice, a model for dry age related macular degeneration. We investigated average axial profiles and compared intensity, spectroscopic, and polarization metrics in the choroid. We found statistically significant differences between groups, suggesting pigmentation changes within the choroid. Analysis of axial profiles reduces the complexity and of analysis which may be valuable in a clinical setting for preventative care, where specific features are not necessarily expected, but trends may be observed over time.

A comprehensive assessment of the retina OCT images is essential in clinical and experimental ophthalmology for the detection and accurate interpretation of the structural and functional abnormalities of the retina. An accurate retinal layer segmentation will allow the measurement of the local and global changes in retina. Herein, we present a computational framework based on a fast-automatic graph-cut based algorithm for an accurate segmentation of the retinal layers for multiple species including human, non-human primate (rhesus macaque) and mouse eye. The proposed algorithm allowed the simultaneous extraction of the multiple biomarkers including retinal layer thickness maps, retinal layer reflectivity maps and layer-specific vascular maps.

It has been recently demonstrated that structures corresponding to the cell bodies of highly transparent cells in the retinal ganglion cell layer could be visualized noninvasively both in the living human and mouse eyes by optical coherence tomography (OCT) via temporal averaging. Here, we further explored the application of volumetric temporal averaging in mice, with a focus on correlating the in vivo results with the ex vivo histology, on the same retinas, to verify the structures seen in the in vivo images, which will help to better understand the pathophysiology of these cells.

A simultaneous multimodality imaging system combining photoacoustic microscopy (PAM), optical coherence tomography (OCT), and fluorescence microscopy (FM) was developed, and its performance and safety for ocular imaging was validated on rabbit eye in vivo. Angiogenesis and bleeding can be selectively shown by label free PAM with laser fluence at only 1% of the ANSI safety limit. OCT imaging allows for visualization of the different retinal layers with high axial resolution. The leakage of neovascularization can be verified with FM. The safety of PAM was evaluated by comprehensive longitudinal follow-up with ocular imaging methods, electrophysiology, histology, immunohistochemistry, and electron microscopy.

In an effort to meet the need of imaging technology that reliably measure the dynamics of accommodation, we developed a swept source OCT system for imaging and biometry of the crystalline lens dynamics at high sampling rates (~100 Hz). Preliminary data suggests that high sampling rates improve detection of lens micro-fluctuations during accommodation and that sampling rates higher than 40 Hz might be needed to fully capture crystalline lens dynamics. Long term, the information acquired with the system will improve our understanding of the mechanism of accommodation and enable to design and evaluate new procedures to restore accommodation.

In this preclinical study we demonstrated that treatment with repetitive magnetic stimulation (RMS) protects the epithelial layer in a rabbit model of exposure keratopathy, preventing loss of the epithelial cells and maintaining the function of the corneal barrier under extreme desiccation conditions. A single 15 minutes RMS treatment supported corneal barrier integrity under acute desiccation for 3 months. AS-OCT imaging and histopathology analysis demonstrated the safety of RMS treatment. Our study suggests that repetitive magnetic stimulation may present a novel treatment for protection of corneal epithelium in patients with DES, and may alleviate DES symptoms.

Swept source OCT was used to image crystalline lens of 100 eyes (age range: 9-78 years) and 3-dimensional lens suture structure was visualized in vivo for the first time. Lens suture patterns were extracted using average intensity projections (AIP) or cortical layers of crystalline lens. Our imaging system has capacity to extract complex star-sutures from cortical layers of lens and simple Y-sutures from fetal nucleus of crystalline lens. Age-related changes in lens and lens sutures were observed and were characterized quantitatively. The developed imaging system can be used to study growth of crystalline lens and its age-related diseases like cataract, presbyopia.

In this presentation we will describe a new method for non-contact in vivo corneal and lenticular microscopy. The technique is based on fundus retroillumination, i.e. anterior segment back-illumination via reflection from the posterior fundus. As such, the retroillumination microscope provides a unique transmission imaging configuration sensitive to forward-scattered light. To enhance intrinsic phase-gradient contrast, we apply asymmetric illumination. The technique produces micron-scale lateral resolution images across a large 1 mm diagonal field of view in the central cornea. We will show representative images of the epithelium, the subbasal nerve plexus, large stromal nerves, dendritic immune cells, endothelial nuclei, and the anterior crystalline lens fibers and nuclei. Finally, we will discuss potential clinical applications and extension to three-dimensional imaging.

The morphology and histology of the corneal nerves in health and disease has been extensively studied, however, the function of these nerves has only been studied in a limited fashion with electrophysiology of the ciliary nerves or calcium reporter dyes in ex vivo corneas. Here, we present non-contact methods for imaging the genetically encoded calcium indicator GCaMP6f in murine corneal nerves both in situ in ex vivo globes and in vivo. These tools have great potential to improve understanding of diseases of the corneal nerves and the development of therapies for diminished neural function.

Optical coherence tomography (OCT) is an interferometric imaging technique that provides non-destructive volumetric visualization of tissues. OCT has been previously proposed as a method to analyze lens cataracts in humans and animal models. In our previous work we showcased the use of a custom-made OCT system to image the crystalline lenses of mice in vivo based on scattering contrast. We now extend our previous results and propose the attenuation coefficient as a parameter for quantitative mapping the lenticular opacities, revealing a significant difference between eyes with ocular cataracts and normal eyes.

Photovoltaic retinal prosthesis is designed to restore sight in patients who lost central vision due to atrophic AMD. Subretinal pixels convert pulsed NIR light projected from augmented-reality glasses into electric current, stimulating the nearby inner retinal neurons. In patients with geographic atrophy, such prosthetic central vision coexists with natural peripheral sight, and its acuity closely matches the 100um pixel pitch of the implant. We present a progress toward 20um pixels based on honeycomb configuration of the stimulating arrays with return electrodes elevated on vertical walls, designed to leverage retinal migration for decoupling the stimulation threshold from pixel size.

Glaucoma causes progressive loss and degeneration of retinal ganglion cell (RGC) axons and their somas. Until recently, cellular-level RGC imaging in live human subjects was not possible. Our initial cross-sectional study revealed differences in morphological parameters in ganglion cell layer (GCL) soma with glaucoma. The manner in which cellular-level morphology is altered during glaucoma progression has yet to be discovered. In this study, we used AO-OCT to track RGC morphology in one glaucoma patient with a hemifield defect. We found dynamic tissue remodeling in the GCL, resulting in a decrease in soma density and an increase in soma size.

The development and application of adaptive optics (AO) in retinal imaging have enabled visualization of a plethora of retinal cells and structures. However, major hurdles exist for translating these achievements to the widely-available clinical devices for broad clinical applications. Here, by configuring a research grade AO – optical coherence tomography (AO-OCT) system to simulate a clinical OCT device, we provide evidence that clinical OCT systems have the potential to resolve individual ganglion cell layer somas and determine that a lateral sampling of ~1.5 µm/pixel is required to accurately quantify soma density and size.

The treatment of macular diseases requires frequent monitoring by optical coherence tomography (OCT). Home monitoring would reduce the burden of frequent clinical visits and increase therapy adherence. In a pilot study with 47 patients having different macular diseases we tested a proprietary self-examination low-cost full-field OCT (SELFF-OCT). For comparison, scans with a standard clinical spectral domain OCT were taken. Data was graded by a reading center. Patients were able to successfully acquire images that were clinically gradable for 85% of the included eyes. The sensitivity and specificity for an anti-VEGF treatment decision based on the SELFF-OCT was 0.94 and 0.95, respectively.

Retinal laser photocoagulation (LP) is commonly used to treat Diabetic Macular Edema. Yet, it is impossible with current lasers to prevent some degree of damage to healthy surrounding tissues. We evaluate through simulation the improvement in thermal confinement of the photocoagulation effects, using Adaptive Optics (AO), a technology that corrects eye’s aberrations in real-time, to assist LP system. Based on experimental ocular aberrations data and realistic retinal diffusion model, we simulate the photocoagulation operation and compare LP thermal confinement in laser systems without and with AO correction. Resulting sizes of thermal damage show an improvement of depth confinement, from a 300μm extension with current systems, to a 70μm depth size (limited to therapeutic target) with AO corrected system. These results underline the need of AO for focal LP to ensure thermal depth confinement and guarantee safe operation.

To evaluate the clinical utility of Visible and near infrared optical coherence tomography (vnOCT) for glaucoma early detection, A total of 55 eyes from three groups of subjects (normal subjects, glaucoma suspects, glaucoma patients) were scanned by a custom-designed vnOCT device and Zeiss Spectralis OCT. The peripapillary retinal nerve fiber layer reflectivity in visible light OCT and the ratio between visible and NIR channel is more sensitive in separating suspect eyes from normal ones than clinical OCT thickness measurements. It could be a useful metric in early detection of glaucoma upon further longitudinally validation.

Human color vision depends fundamentally on three spectral types of cone photoreceptors, yet methods to objectively measure these types across the whole visible spectrum in the living human eye do not exist. Here we demonstrate a new method based on phase-sensitive adaptive optics optical coherence tomography that offers the sensitivity and resolution to obtain spectral sensitivities with extremely high precision from 450-635 nm. We present the first objective measurements of human cone spectral sensitivity from 450 – 635 nm for all three cone types and demonstrate a path for quantifying spectral sensitivity over the whole visible spectrum.

Phase-sensitive optical coherence tomography (OCT) is emerging as an imaging modality that detects functional changes in the retina. Besides imaging photoreceptor function, recently, functional changes in the inner plexiform layer (IPL) have been detected using full-field swept-source OCT. The IPL connects neuronal cells which are dedicated for processing different aspects of the visual information, such as edges in the image or temporal changes. A characteristic of signal processing in the IPL is that different aspects of the visual impression are only processed in very specific depths. Here, we present an investigation of these functional signals for different depths in the IPL with the aim to separate different properties of the visual signal processing. Therefore, we investigate the phase changes of three different sub-layers. Whereas the first two depths, closest to the ganglion cell layer, exhibit an increase in the optical path length, the third depth, closest to the bipolar cell layer, exhibits a decrease in the optical path length. Additionally, we found that the second or middle depth is sensitive to temporal changes, showing a maximum increase of the optical path length at a stimulation frequency of around 10 Hz. The results suggest that the responses from different cell types, which are sensitive to different features of the stimulation signal, can be distinguished by phase-sensitive OCT.

The development of functional retinal imaging is of great interest to clinical and experimental ophthalmology. Recent progress in OCT-based functional imaging of rods and/or cones outer segments elongations in response to stimuli that bleach fraction of their rhodopsin promises to address the need for probing photoreceptor function. Building on this model, we developed the framework to monitor changes in the gross morphology of the whole retina of mice in response to light stimulation. In our current submission, we will discuss our findings in the context of known retinal water barriers restricting water movements between inner and outer retina.

Early detection of photoreceptor dysfunction is essential for preventing vision loss due to retinal degenerative diseases, such as age-related macular degeneration (AMD) and inherited retinal degenerations (IRDs). Functional intrinsic optical signal (IOS) imaging promises a high-resolution method for objective optoretinography (ORG). Stimulus-evoked photoreceptor-IOS has been recently demonstrated in healthy animal and human retinas. The fast photoreceptor-IOS response was found to occur at the photoreceptor outer segment (OS) right after the onset of retinal stimulation. However, in vivo IOS response of photoreceptor dysfunctions is not yet validated, which is essential to measure the clinical usability of ORG measurement. In this study, we report in vivo IOS imaging of rod photoreceptor dysfunction in retinal degeneration 10 (rd10) mice. A custom-designed optical coherence tomography (OCT) was used for photoreceptor-IOS imaging. A significant attenuation of the photoreceptor-IOS was found in rd10 mice due to disorganized ultrastructure of the photoreceptor OSs, which appeared ahead of progressive rod cell death. Our experiments demonstrate that fast photoreceptor-IOS is highly sensitive to ultrastructural integrity of the photoreceptor OSs. We anticipate that quantitative imaging of fast photoreceptor-IOS will provide objective ORG measurement to advance the study and diagnosis of AMD, IRDs, and other retinal diseases that can cause photoreceptor dysfunctions.

This study is to characterize intrinsic optical signal (IOS) changes in the photoreceptor outer segment (OS) and inner segment (IS). Functional optical coherence tomography (OCT) was used for in vivo IOS imaging of wild-type mice (C57BL/6J). Depth-resolved OCT revealed stimulus-evoked IOSs at individual retinal layers, with sub-photoreceptor resolution. Rapid IOS response occurred in the OS region immediately after the stimulus onset. In contrast, transient IOS showed a time delay and gradually increased in the IS region. We anticipate that the OS-IOS and IS-IOS can work as objective biomarkers to reflect photoreceptor phototransduction and metabolism property, respectively.

Physiological dysfunction of diseased cells might occur prior to detectable morphological abnormalities such as retinal cell damage and thickness change. Functional assessment of photoreceptor physiology is essential for the early detection of eye diseases. It is desirable to develop a high-resolution method for objective assessment of photoreceptor physiology. Functional intrinsic optical signal (IOS) imaging, also known as optoretinography (ORG) or optophysiology, measures transient light changes correlated with retinal neural activities. The photoreceptor-IOS arises promptly after the beginning of stimulation, which ensures a unique biomarker for objective ORG measurement of physiological conditions of retinal photoreceptors. In this study, the feasibility of functional optical coherence tomography (OCT) imaging of fast photoreceptor-IOS in human photoreceptors has been demonstrated. The fast photoreceptor-IOS occurred before stimulus-evoked pupillary response and thus allows nonmydriatic ORG of human photoreceptors. The outer segment (OS) was confirmed as the source of fast photoreceptor-IOS by depth-resolved OCT. The active IOS changes were found at both OS boundaries, which connected to the inner segment and retinal pigment epithelium. This supports that the mechanism of the fast photoreceptor-IOS can be explained by transient OS shrinkage due to phototransduction.

Increasing the FOV of OCTA images while keeping the acquisition time moderate requires high A-scan rates. Therefore, OCTA images appear to be noisier. Deep learning methods can be used for noise reduction. In OCTA volumes small vessels with an orientation perpendicular to the image plane are often removed by deep learning denoising algorithms, due to their small appearance. To overcome this a 3-dimensional Unet was developed to utilize volumetric information. With the knowledge of also the third dimension, the algorithm is able to distinguish between noise and vessel contrast and is therefore less likely to remove vessels.

Conventional optical coherence tomography (OCT) is unable to visualize the earliest microstructural changes that accompany retinal dysfunction and pathology, delaying diagnoses in diseases such as glaucoma. We present methods for analyzing light scattering behavior using OCT to report microscopic structural changes using our layer-based, depth-resolved attenuation coefficient, and layer-resolved backscattering fraction. These quantitative scattering parameters are sensitive to sub-resolution status of pathologically relevant features such as constituent cells, intracellular components, and fiber organization. Preliminary results show additional feature contrast in a healthy subject, and we look forward to exploring the future clinical potential of these methods to enable earlier diagnoses.

Early disease diagnosis and effective treatment assessment are crucial to prevent vision loss. Retinal arteries and veins can be affected differently by different eye diseases, e.g., arterial narrowing and venous beading in diabetic retinopathy (DR). Therefore, differential artery-vein (AV) analysis can provide valuable information for early disease detection and better stage classification. However, manual, or semi-automated methods for AV identification are inefficient in a clinical setting. This study is to demonstrate the use of deep learning for automated AV classification in optical coherence tomography angiography (OCTA). We present ‘AV-Net’, a fully convolutional network (CNN) based on a modified Ushaped architecture. The input to AV-Net is a 2-channel system that combines grayscale enface OCT and OCTA. The enface OCT is a near infrared image, equivalent to a fundus image, which provides the vessel intensity profiles. In contrast, the OCTA contains the information of the blood flow strength, and vessel geometric features. The output of AV-Net is an RGB (red-green-blue) image, with R and B corresponding to arteries and veins, respectively, and the G channel represents the background. The dataset in this study is comprised of images from 50 individuals (20 controls and 30 DR patients). Transfer learning and regularization techniques, such as data augmentation and cross validation, were employed during training to prevent overfitting. The results reveal robust vessel segmentation and AV classification. A fully automated platform is essential for fostering efficient clinical deployment of AI-based screening, diagnosis, and treatment evaluation.

Cataract surgery is the most frequently performed surgical procedure in all of Ophthalmology. During this process of surgically removing the lens the inner layer of the cornea needs for protection from permanent damage and the anterior chamber demands stabilization, which is provided by Ophthalmic Viscosurgical Devices (OVDs). We therefore present an automatic pipeline to determine the thickness of OVDs in enucleated porcine eyes via semantic segmentation. For the evaluation of the pipeline we segmented 100 volume scans of prepared porcine eyes. This versatile pipeline can be applied for segmentation of anterior segment Optical Coherence Tomography scans of the anterior segment.

The present project aims at developing a fully automatic software for estimation of the waist of the nerve fiber layer in the Optic Nerve Head (ONH) angularly resolved in the frontal plane as a tool for morphometric monitoring of glaucoma. The waist of the nerve fiber layer is here defined as Pigment epithelium central limit –Inner limit of the retina – Minimal Distance, (PIMD). 3D representations of the ONH were collected with high resolution OCT in young not glaucomatous eyes and glaucomatous eyes. An improved tool for manual annotation was developed in Python. This tool was found user friendly and to provide sufficiently precise manual annotation. PIMD was automatically estimated with a software consisting of one AI model for detection of the inner limit of the retina and another AI model for localization of the Optic nerve head Pigment epithelium Central limit (OPCL). In the current project, the AI model for OPCL localization was retrained with new data manually annotated with the improved tool for manual annotation both in not glaucomatous eyes and in glaucomatous eyes. Finally, automatic annotations were compared to 3 annotations made by 3 independent annotators in an independent subset of both the not glaucomatous and the glaucomatous eyes. It was found that the fully automatic estimation of PIMD-angle overlapped the 3 manual annotators with small variation among the manual annotators. Considering interobserver variation, the improved tool for manual annotation provided less variation than our original annotation tool in not glaucomatous eyes suggesting that variation in glaucomatous eyes is due to variable pathological anatomy, difficult to annotate without error. The small relative variation in relation to the substantial overall loss of PIMD in the glaucomatous eyes compared to the not glaucomatous eyes suggests that our software for fully automatic estimation of PIMD-angle can now be implemented clinically for monitoring of glaucoma progression.

The physical distancing requirements necessary to prevent spread of the novel coronavirus, SARS-CoV-2, requires a change in approach for clinical ophthalmic imaging. Conventional optical coherence tomography (OCT) systems require patients to position themselves in chin/forehead rests for stabilization with the system operator in close proximity. We developed a robotically aligning OCT (RAOCT) system that provides volumetric retinal images encompassing both the optic nerve head and fovea. Our RAOCT system self aligned to subjects’ eyes (seated, no contact with restraints), acquired OCT images of both normal and diseased retinas, all with allowing the operator behind a barrier >2 m from the subjects.

Bruch’s membrane (BM) is a pentalaminar structure that mediates transport between the retinal pigment epithelium (RPE) and choriocapillaris. With near-infrared Optical Coherence Tomography (OCT), it has been challenging to visualize, let alone quantify, BM non-invasively in non-pathologic eyes. First, we show that shorter wavelength visible light OCT consistently delineates BM better than longer wavelength visible light OCT in pigmented human subjects, independent of axial resolution. Second, we develop a physical model of RPE and BM reflectivity to explain this finding. Third, we employ this model to devise a morphometric algorithm to more accurately map BM thickness in the normal macula.

High-resolution ophthalmic imaging is imperative for detecting subtle changes of photoreceptor abnormality at the early stage of retinal diseases. However, optical resolution in retinal imaging is inherently limited by the low numerical aperture of the ocular optics. Virtually structured detection (VSD) has been demonstrated to break the diffraction limit of imaging systems by shifting the high-frequency components to the passing bandwidth of the imaging system. However, its implementation for human subjects remains a challenge due to the uncertain cut-off frequency of the modulation transfer function (MTF) required for VSD processing. This study demonstrates an objective method to derive the MTF from spectral profiles, enabling quantitative estimation of the optimal cut-off frequency. A custom-built line-scan scanning laser ophthalmoscopy was developed, and two-dimensional line-profile patterns were acquired at a 25 kHz frame rate. We found that the MTF profiles exhibited significant differences between subjects as well as view fields. VSD-based super-resolution images exhibited improved resolution and contrast to differentiate individual photoreceptors compared to the equivalent wide-field imaging. Besides, the motility process on the VSD image further improved the image quality as the photoreceptors revealed clear boundaries and more integrated shape, compared to that in the VSD image. We anticipate that the VSD-based imaging will provide a simple, low-cost, and phase-artifact-free strategy to achieve super-resolution retinal ophthalmoscopy.

Fourier-domain full-field optical coherence tomography (FD-FF-OCT) is currently the fastest volumetric imaging technique that is able to generate a single 3D data volume in less than 10 ms. FD-FF-OCT is based on a fast camera, a rapidly tunable laser source and a Fourier-domain signal detection. However, crosstalk corrupts images with speckle and therefore lowers image quality. Corneal imaging with FD-FF-OCT is also challenging since the laser focuses down on the retina causing safety concerns. We present spatially incoherent FD-FF-OCT system that reduces crosstalk noise in retina images and enables corneal imaging by reducing spatial coherence of the laser.

Increasing the field of view (FOV) of optical coherence tomography (OCT) for the high-resolution posterior eye imaging with a continuous scan is demonstrated. The combination of a Lissajous trajectory, which is designed for high-resolution imaging, and a slow drift was applied to scan the probing beam. The FOV is increased as the slow drift progresses. The motion artifacts are suppressed by motion estimation and correction in post-processing. The high-resolution, motion-free OCT and OCT angiography imaging with a FOV of over 6.75 mm is achieved.

Confocal Scanning Laser Ophthalmoscope (cSLO) is a retinal imaging technique based on the principle of confocal microscopy which provides high-contrast and high-resolution fundus eye imaging. Acousto-optic lenses (AOL) are tunable lenses enabling fast focus tuning that can be applied in rapid depth scanning. We propose a cSLO integrated with AOL to generate enhanced in vivo retinal images. We present the performance of the optical system based on the tests with eye phantom. The measurements included also determination of the refractive error (spherical defocus).

The eye motion is broadly considered as a valuable source of information in the fields related to psychology, neuroscience and neurology. Therefore, quantitative characterization of eye tracker data is an important task. Saccades are sigmoidal, ballistic movements that in particular deserve more attention due to their complex shape and natural diversity. We have developed the high accuracy model of saccades and microsaccades of mean absolute error equal to 0.0104 degree on average. We present the methodology for extraction of saccadic features using this model and show the potential of the method in biometric experiments.

Mirrors in resonant galvanometric optical scanners experience dynamic distortion due to torque which is proportional to angular displacement. The dominant aberration introduced by the sinusoidal oscillation of the scanner is oblique astigmatism that is linear with the field coordinate. Nodal aberration theory (NAT) indicates that such linear astigmatism can be compensated by tilting elements with power in an optical system. Here, we demonstrate a practical and low-cost solution that only requires tilting of existing optical elements immediately following the resonant scanner. The proposed method can be used to generate any desired third-order aberration that results from tilting or decentering optical surfaces in optical systems. This is illustrated by correcting linear astigmatism due to a resonant scanner by tilting elements in reflective and refractive afocal relays. In all cases, wavefront correction better than the classical diffraction limit (Strehl ratio > 0.8) was demonstrated.

Pulsed near-infrared (NIR) light sources can be successfully applied for both imaging and functional testing of the human eye, as published recently. These two groups of applications have different requirements. For imaging applications, the most preferable is invisible scanning beam while efficiently visible stimulating beam is preferable for functional testing applications. The functional testing of human eye using NIR laser beams is possible due to two-photon vision (2PV) phenomenon. 2PV enables perception of pulsed near-infrared laser light as color corresponding to approximately half of the laser wavelength. This study aims to characterize two-photon vision thresholds for various pulse lengths from a solidstate sub-picosecond laser (λc = 1043.3 nm, F_rep = 62.65 MHz), either of 253 fs duration or elongated by Martinez- type stretcher to 2 ps, and fiber-optic picosecond laser (λc = 1028.4 nm, F_rep = 19.19 MHz, τ_p = 12.2 ps).

This study is to validate the trans-pars-planar illumination for ultra-widefield multispectral imaging (MSI) of the retina and choroid. By freeing the available pupil for collecting imaging light only, the trans-pars-planar illumination enables a portable, nonmydriatic fundus camera, with 200o FOV in a single-shot image. The trans-pars-planar illumination, delivering illumination light from one side of the eye, naturally enables oblique illumination ophthalmoscopy to enhance the contrast of fundus imaging. Four wavelength LEDs, including 530 nm, 625 nm, 780 nm, and 970 nm, are illuminated for MSI of the retina and choroid.

Diabetic retinopathy (DR) is a leading cause of preventable blindness. Early detection and reliable stage classification are essential to ensure prompt medical interventions. Recent study suggests that the outer retina, i.e., photoreceptors, can be affected by early DR. We demonstrate here the potential of using quantitative OCT features in outer retina for objective detection and stage classification of DR. The OCT intensity change is observed to be mostly sensitive, compared to retina thickness and bandwidth, to DR stages. It is also confirmed that the relative intensity changes of photoreceptor outer segment are more sensitive than inner segment for DR classification.

Off-axis full-field OCT is intended to enable cost-effective imaging of the retina for home diagnosis. Different to common FD-OCT systems, the lateral field of view is acquired in a single shot and the different axial layers are acquired sequentially. During acquisition, motion of the eye results in motion artifacts and misaligned layers. We present a method to track the axial and lateral position of the retina by analyzing the angle and divergence of the backscattered light with a lateral precision of 3.6 µm and an axial precision of 29 µm. This information can be used to correct motion induced errors.

Changes in the protein aggregation within the ocular lens may be responsible for both presbyopia and cataracts. Treatments for these conditions are under development, but are likely to be most effective when administered early in the disease. Therefore, a technique compatible with in vivo use which could detect early changes in aggregations is desired. Here, we assess if phase-decorrelation OCT may be sensitive to cold-induced protein aggregation in ex vivo porcine lenses. A major challenge of this approach is the relatively weak scattering signal obtained from the lens nucleus while the lens is in situ. We observed a substantial increase in decorrelation time as the cold cataract reversed. Backscatter intensity also decreased as the cold cataract reversed, as expected. However, compared to backscatter intensity, decorrelation is better correlated with cataract reversal.

Constructing an image acquired by a non-uniform scanning pattern is a difficult task. The main challenges are:(1) resampling technique (2) discrepancy between demanded (dictated by control signal) and actually performed, empirical scanning path. Here, we show how to calibrate the scanning path of MEMS scanner using Galvanometric Scanner and to what extent the time of acquisition impacts the resulting image.

Optical coherence tomography angiography (OCTA) is a well-established retinal imaging modality that is emerging as a fast, non-invasive alternative to fluorescence angiography for assessment of corneal injury and neovascularization caused by chemical injuries, infections, and other sources of corneal damage. OCTA algorithms typically perform operations on multiple scans, or frames, at the same location to identify flowing vasculature. In this work, we describe a novel singleframe algorithm that relies on common image processing operations, allowing for broad application to various OCT systems, as well as reduced acquisition and computation times. We also show the potential of a multi-frame approach, based on the same principle, that allows for enhanced discrimination between flowing and static anatomical features. To demonstrate the capability of our approach, we processed the same image stack with our single-frame and multi-frame algorithms along with other angiography algorithms, such as phase variance, speckle variance, and complex differential variance and found that our algorithms had higher estimated signal-to-noise ratios (SNR) and lower computation times. We applied our algorithms to quantifying corneal neovascularization (CoNV) in a murine model of corneal burn injury through semi-automated measurement of vessel area and compared them to the gold standard of fluorescein angiography. This work provides strong evidence for the power of the single-frame algorithm and its multi-frame variant, as well as the potential of OCTA for quantification of corneal pathology beyond the standard fluorescein angiography approach allowing for more accurate monitoring and staging of corneal injury and wound healing.

Photo-mediated ultrasound therapy (PUT) holds potential as a novel antivascular method. In this work, we applied PUT to precisely remove corneal neovascularization in a rabbit eye model. A stable corneal suture-induced corneal neovascularization model was established in rabbits. These rabbits were later treated by PUT or used as controls. The treatment outcomes were evaluated through red-free photography and fluorescein angiography along with histology and immunohistochemistry. The experimental results demonstrated that PUT shows promise in improving the management of eye diseases by delivering selective treatment to pathologic vessels with minimized side effects.

Pupillary light re ex involves many sensory and motor functions of the eye. For this reason, it represents an important emergency diagnostic tool and provides information to assess brain stem function. The pupil system can be considered in terms of input-output black-box behavior: light stimuli can be easily applied to the eyes, and the pupil size can be measured effortlessly and non-invasively. In this paper, a model for short-light- ash-induced transient pupillary light re ex is presented and preliminary experiments designed to test the model features are described. Results confirm that the developed pupillary light re ex model is suitable to describe the pupil oculomotor system exposed to short-light- ashes.

Primary Angle Closure Glaucoma occurs more frequently in people with a narrower limbal anterior chamber depth (LACD) condition. Nowadays, clinical gold standard as an assessment technique, i.e. gonioscopy, is invasive and complex, whereas Van Herick (VH) technique is non-invasive, but subjective. The instrument, we propose, can automatically performs the VH procedure using a blue laser line, a piezo-actuator, and an image recognition algorithm embedded on a Raspberry Pi board. Preliminary measurements have been carried out on volunteers, and the results proved the feasibility of our approach. The final instrument unveils a high potential for early-stage diagnosis and screening applications.

Retinal vein occlusion (RVO) is a common cause of vision impairment and blindness. In RVO, thrombosis in a retinal vein and reduction in blood flow leads to the development of retinal hypoxia. Retinal hypoxia stimulates vascular endothelial cell growth factor (VEGF) production, resulting in the formation of neovascularization and vision loss. Currently, non-invasive imaging modalities are not available to track and quantify the tissue-level hypoxia in living animals. In this study, a novel organic room-temperature phosphorescence nanoparticle (PNPs) was developed and characterized for identification of retinal hypoxia in real-time in living rabbits. The ability of PNPs for hypoxia imaging was examined in rabbits with laser-induced RVO models (n=3). The location of retinal hypoxia was imaged before and after the intravitreal administration of PNPs at a concentration of 2.5 mg/mL and an injection dose of 50 µL. Phosphorescence imaging was acquired at different time points over 7 days alo

Retinal vasculature is affected in many ocular conditions including diabetic retinopathy, glaucoma and age-related macular degeneration and these alterations can be used as biomarkers. Therefore, it is important to segment and quantify retinal blood vessel characteristics (RBVC) accurately. Using a new automated image processing method applied to optical coherence tomography angiography (OCTA) images we computed the RBVC and compared them between emmetropic (n=40) and ametropic (n=97) subjects. All 137 OCTA images had dimensions of 420x420 pixels corresponding to 6mm x 6mm. The myopia OCTA images were labelled based on a severity scale as mild, moderate, high and very high using standard refractive error classifications. Before image processing, all the images were cropped to 210 X 210 pixels keeping the foveal avascular zone (FAZ) at the centre to quantify the RBVC. The mean ± standard deviation of the Grisan index, a measure of retinal blood vessel tortuosity in the emmetropic, and myopic eye were 0.05 ± 0.02 and 0.05 ± 0.03 respectively. The total vessel distance measures were calculated and the largest were found in emmetropic eyes (45.95 ± 19.54) and shortest in myopic eyes (6.50 ± 5.17). The total number of turning points and inflection points were found to be statistically significant (p<0.05) between control and myopic eyes. However, other RBVC parameters were not statistically different (p=<0.05). We found qualitatively that RBVC changes with increasing severity of the refractive power. Among RBVC parameters, average number of turning points (NTP) decreasing trend with degree of myopia increases.

Uniform and quantitative grading of retinal vessel characteristics are replacing subjective and qualitative schemes. However clinically accurate blood vessel extraction is very important. The tortuosity of these vessels is an important metric to study the curvature variations in normal and diseased eyes. In this study we provide a new unsupervised and fully automated approach for studying curvature variation of the blood vessels. We then pro- vide tortuosity quantification of these extracted vessels. In this study we used optical coherence tomography angiographic fundus images of dimensions 420x420 pixels corresponding to 6mm x 6mm were used in this study. We focused on the central circular 210x210 pixel region around the foveal avascular zone (FAZ) for tortuosity quantification. Our segmentation approach starts with a 3mm x 3mm central circular region extraction surrounding the FAZ. We then use a multi-scale, multi-span line detection filter to smoothen out the high noise in the background and at the same time increase the intensity of target vessels. This is followed by a K-means procedure to filter out the noise and target vessels into two categories. Next steps are morphological closing and noise removal and iterative erosion of pixels to skeletonize the vessels. The final extracted vessels are of the form of single pixel piecewise continuous fragments. These are finer than human annotations and at the same time free of noise. We then provide accurate standard tortuosity measures - Distance Measure, Inflection Points, Turning Points, etc. for these OCTA images using the extracted vessels through mathematical modelling.

Different wavelength lasers are widely used in ophthalmology for example for selectively heating certain tissues of the eye or unleashing the potential of photoactive pharmaceuticals. The problem with many ophthalmic laser-based treatments such as photodynamic therapy for age-related macular degeneration is that the laser technology is outdated and no longer supported despite the wide clinical use of these therapy modalities. Modulight has developed a configurable cloud-connected ophthalmic laser device that can house any of Modulight’s semiconductor lasers and is wirelessly controlled with an iPad. In addition to novel ophthalmic laser technology, Modulight has also developed a novel beam shaping unit which yields superior beam quality and enables exceptionally large treatment spots eliminating the need for multi-spot treatment for larger lesions.

The retina regeneration process has been observed in several animals including fish, birds, and amphibians, whereas the injured human retina cannot regenerate until today. The zebrafish is one of the animals which can regenerate their retina. Due to similarities between humans and the zebrafish in the morphology of the eye as well as the gene, the zebrafish has been chosen as a useful model for investigating retina regeneration. Previous studies have observed morphological changes in the zebrafish retina using optical coherence tomography during the retinal regeneration after light irradiation with a single beam power. However, these studies have limitations of demonstrating differences in regenerative abilities depending on degrees of retinal damages because they used only a single light irradiating condition. Through light-induced retinal injuries by various light irradiating conditions, investigation correlations between the time required for retina regeneration and grades of retinal damages were needed. To conduct these experiments, a custom-built OCT, which can acquire a cross-sectional and three-dimensional image of the zebrafish's eye, was developed based on OpticStudio simulations. The zebrafish's eye imaging can provide the ability to observe the damage and morphological changes after laser irradiation in the eye tissues including the cornea and retina.

Cataract is one of the common causes of visual impairment due to opacification of the crystalline lens. In the proposed study, the effect of cataract is simulated and optically corrected in a single-pass system with digital micromirror device (DMD). The correction was carried out by implementing binary amplitude mask which mimics the opacities of the crystalline lens in such a way that the light tends to pass through the transparent parts and there was no light interacting with opacities of cataractous lens. The comparative analysis of the simulated retinal images demonstrates improvement in image quality.

Optic disc tilt (ODT), peripapillary atrophy (PPA), and abnormally large or small optic discs are the earliest known changes in myopic eyes and may precede the development of pathological myopia. Increasing ODT and distance between the macula and optic nerve head have been reported as being associated with progressive myopia. Therefore, it is important to segment and quantify the ODT accurately. Using a newly developed automated image processing method we measured the ODT in both myopes and emmetropes. We determined the ODT from myopic eyes (n= 90) and compared the results with emmetropia (n=14). All 104 optical coherence tomography (OCT) images had dimensions of 200x200 pixels corresponding to 6mm x 6mm. The myopic OCT images were labeled based on a severity scale based on the spherical error (SE) as low (-0.5 to -3.00 D SE), moderate (-3.12 to -6.00 D SE), high (-6.12 to -9.00 D SE), and very high (worse than -9.00 D SE) using standard myopia classifications. Each OCT image was segmented by a clinician (CEM) and by the newly proposed method (NAM). The NAM used 8-bit grayscale OCT images which was preprocessed by applying Gaussian blur and Contrast Limited Adaptive Histogram Equalization based thresholding to remove noise and locate regions of interest. Then the images were split into two halves to fit straight lines separately. Morphological erosion and dilation were performed on the images to remove artifacts. They were tested with three combinations of erosion and dilation iterations. Univariate linear regression Lines were fit to trace the white band on each half and angle between the lines was determined. Both the methods showed higher horizontal ODT than vertical. The mean ± SD horizontal ODT (in degrees) in the myopic eye was 18.47 ± 7.67 and 15.84± 6.61 by the NAM and CEM methods. The vertical ODT in the myopic eyes (in degrees) was 16.32 ± 7.10, and 14.52 ± 7.05 by the NAM and CEM methods respectively. However, the NAM showed a maximum difference (2.26 ± 5.68) between horizontal and vertical ODT. The study results show that the ODT in very high myopic eyes (26.33±8.99)is significantly different (p<0.05) when compared to emmetropic eyes (19.47±3.99).

Traditional refraction measurement equipment is expensive, cumbersome and requires professional training for its operation. Therefore, refraction measurements are usually performed by eye care professionals, requiring an in-office visit. A handheld, low-cost and simple refraction measurement device based on the reverse Shack Hartmann technology is aimed for consumers, telehealth and at-home measurements. The device attaches to a smartphone where patterns on the screen are aligned by the user looking through the optics of the device. The optical train consists of chromatic separation into two channels that then combine in the user’s eye. The measurement consists of repeated alignment of the patterns by the user at different meridians of the pupil to allow for refraction mapping. The formulation of a 2D pattern and alignment scheme is presented for achieving optometric refraction values. In this case, the method maintains a low number of required measurements by the user. On the other hand, higher order aberrations may be measured using this method and the general terms of the measurement formulation is presented using the Zernike polynomials. The derivation is based on a model of the eye with an angular deviation field as the source of refraction. The paraxial approximation is used as the angular deviations are shown to be small for any ophthalmic condition. The model and formulation are then validated by a series of experiments in a calibrated model eye. Results show that two measurements are sufficient for producing accurate optometric refraction measurement values.