Ophthalmic Technologies XXXV

In the current study the age related loss rate of the minimal cross section of the waist of the nerve fiber layer in the optic nerve head (ONH) was estimated with OCT as a reference for glaucoma monitoring. The ONH of both eyes in 44 subjects, evenly distributed in the age interval [30;70] years, with an even gender distribution, were captured with the OCT-Triton (Topcon, Japan). The minimal cross section of waist of the nerve fiber layer in the ONH was estimated fully automated with a custom made software both as thickness (PIMD-2pi) and area PIMA-2pi. A 95 % confidence interval for the age related loss of PIMD was estimated to -1.53 ±1.37 um·yr^-1 (d.f. = 42) and loss of PIMA to -0.44 ±0.9 x10-2 mm^ 2 ·yr^-1 (d.f. = 41).

We developed a compact lens-based multimodal optical coherence tomography (OCT) and scanning laser ophthalmoscopy (SLO) high-resolution system for in vivo mouse retinal imaging that enables the visualization of retinal laminar microstructures, microvasculature, and labeled neuron cells, simultaneously.

Eye movement analysis can provide valuable information about the condition of the central nervous system. However, currently used video-based eye trackers are insufficient to reliably examine eye micro-movements. Therefore we present a retinal-based eye tracker NeuroFET based on an optical arrangement of a confocal scanning microscope optimized for retinal imaging and achieves accuracy up to 0.02° and precision up to 0.0005°. Pilot studies shows that NeuroFET can be used to capture subtle diurnal changes resulting from the subject's psychophysical fatigue and sleep deprivation. This makes it a promising device for monitoring neuronal changes in neurodegenerative and ophthalmological diseases.

We developed a dual channel adaptive optics scanning laser ophthalmoscope (DualCH-AOSLO) for in vivo 3D imaging of microglia in mice. The results showed that all three vascular layers were distinctly visualized by angiography, providing a reference for localizing microglia. AO significantly enhanced imaging quality, allowing clear distinction of the cellular processes and branching patterns within individual microglia. Inactive, more ramified microglia were primarily observed in IPL, INL, and OPL, while fewer microglia were found in the nerve fiber layer (NFL). This system proves to be a valuable tool for in-vivo monitoring of microglial activity.

Multiple sclerosis (MS) is a neurodegenerative disease that impacts up to one million people in the U.S., with females twice as likely to be diagnosed as males. While optical coherence tomography (OCT) has provided significant information on inner retinal layer thickness changes in MS, cellular-level resolution is necessary to fully understand the cause and course of the disease, an essential prerequisite to the development of new treatment strategies. Using clinical OCT and adaptive optics – optical coherence tomography (AO-OCT), we collected volumes in the peripapillary region of MS patients and characterized immune cells and other structures important to the disease.

Glaucoma causes progressive loss of retinal ganglion cell (RGC) function and eventually RGC apoptosis as well as changes in the retinal blood flow (RBF). Understanding the dynamic relationship between RGC function and RBF (defined as neurovascular coupling), can help ophthalmologists to understand better the origins and stages of progression of glaucoma. Here we present a novel combined OCT+ERG system capable of simultaneous imaging of retinal morphology, probing of retinal function (ORG – optoretinography and ERG – electroretinography) and measurement of visually evoked transient RBF changes. The system was used in a rat model of glaucoma to investigate changes in the retinal neurovascular coupling.

Visual stimulation-induced increase in the metabolic activity of retinal neurons leads to temporary vasodilation of retinal blood vessels and an increase in the retinal blood flow, which is often referred to as functional hyperaemia or neurovascular coupling. Glaucoma, a neurodegenerative retinal disease, causes progressive damage to the retinal morphology, neuronal function and blood flow, and eventually leads to blindness. In this study, we utilize a combined OCT+ERG system to investigate neurovascular coupling in the healthy and glaucomatous human retina.

Loss of retinal ganglion cells (RGCs) is a hallmark of glaucoma, with structural damage preceding visual field defects. Detecting RGC loss clinically is challenging due to slow disease progression and lack of sensitive biomarkers. Adaptive optics – optical coherence tomography (AO-OCT) offers noninvasive cellular imaging of glaucomatous damage. This study quantified changes in ganglion cell layer (GCL) soma density and diameter using AO-OCT, comparing glaucomatous eyes to healthy volunteers longitudinally. Accurate cell density measurements may prove to be a sensitive glaucoma biomarker, enhancing early detection and monitoring of disease progression and treatment efficacy

Using two or more vertically shifted B-scans, we present a new method to extract the confocal and fall-off functions as a single function. In OCT, removing the effects of these functions improves the quantification of the attenuation coefficient (AC). This method uses basis functions to achieve a better data-to-model fit of the combined function and does not require prior knowledge of the OCT system parameters or the forms of confocal and fall-off functions, typical of other methods. The method is demonstrated on human retina OCT data to produce the AC images and shows improvement over the standard ratio fit method.

We present an 840 nm swept-source OCT system with an A-Scan rate of 1.7 MHz. The laser shows promising performance for retinal imaging with 10 mW output power and a theoretical axial resolution in water of 12 µm. We describe the technical specifications and performance data and show first imaging results of an eye phantom model. The resolution of the first images looks good, the signal levels and sensitivity promising. We also plan to present data of our in vivo imaging sessions in the near future as well as performance specifications of a 3.4 MHz system.

Understanding how the eye moves is of utmost importance to scientists in neurology, ophthalmology, cognitive sciences, and medicine. Popular eye trackers observe the anterior parts of the eye, but a new class of eye-trackers looks at the retina. Each of the two techniques has advantages and disadvantages: the anterior eye trackers provide data with low precision but high accuracy, while retinal eye trackers provide data with high precision but low accuracy. We have built a device that combines the two trackers and allows for the simultaneous acquisition of eye motion on the anterior and posterior parts of the eye during visual tasks displayed to the subject, such as fixations, saccades, and smooth pursuit. The retinal tracker is based on the concept of SLO, but modified to acquire small image frames (5deg x 3deg) at high speeds. The anterior tracker uses a CMOS camera to register the pupil and the Purkinje reflections.

We present a novel PS-OCT framework that integrates with clinical retinal spectral-domain OCT systems with minimal modifications. The approach involves inserting a rotating achromatic half-wave plate into the sample arm and acquiring repeated B-scans while the waveplate is rotated. Ideally, scatterers in non-birefringent tissues produce identical interference signals across all waveplate positions, while birefringent tissues cause a rotation-dependent modulation. Three polarimetric measurements at distinct waveplate angles are sufficient to determine the round-trip polarization effects of the sample. We demonstrated the cumulative retardance and cumulative optic axis orientation images of a birefringent phantom and healthy eyes using the novel PS-OCT framework.

Optical coherence tomography (OCT) is a critical imaging technique for diagnosing diseases of the eye. Whole-eye OCT acquires images of the cornea and retina simultaneously, allowing biometry of the eye and subsequent quantitative retinal curvature retrieval. However, patient motion during acquisition can corrupt spatial relationships in the final image. Additionally, OCT for the anterior and posterior segments of the eye require distinct instrumentation. Here we present a widefield whole-eye OCT system with integrated pupil tracking cameras for tracking patient motion. Using patient motion data, we reduce patient motion artifacts in our quantitative retinal curvature retrieval algorithm.

We quantify Bruch's membrane (BM) thickness changes with age and eccentricity using visible light Optical Coherence Tomography (OCT) with approximately 1 micrometer axial resolution. BM was distinguished by visible light OCT, with better contrast than near-infrared OCT. Qualitatively, BM gets thicker and less distinguishable in older subjects. To quantify BM thickness, axial scans were aligned and averaged, while retinal pigment epithelium multiple scattering was corrected using a previously-validated procedure. We observed BM thickening with both age and eccentricity. These results suggest that visible light OCT can serve as an in vivo imaging tool to quantify morphological changes in human BM.

Retinal image-based eye tracking systems are important for health and vision science applications requiring arcmin precision. However, current systems are limited in gaze range due to challenges in consistent retinal visualization, often impacted by pupil size, eye movement, and subject misalignment. While pupil-steering can mitigate these problems, previous implementations haven’t significantly expanded gaze range without degrading accuracy and/or latency. This work introduces a novel eye tracker that employs eye center-conjugate optical steering, driven by pupil tracking, to increase the dynamic range of gaze-matched retinal imaging to over a 50° gaze range, capturing eye motion at 200 Hz with arcmin precision.

Cone and rod photoreceptors are the most studied cells in the living human retina using adaptive optics (AO) imaging systems. However, major components of the photoreceptor cell, namely its soma and axon, remain challenging to image due to their high transparency and extremely weak reflections. Imaging photoreceptor somas is of significant clinical interest as the soma is responsible for the cell’s genetic information and central for maintaining cellular health. Here, we demonstrate a method based on AO-OCT imaging that reveals, for the first time, the 3D mosaic of photoreceptor somas in healthy eyes and detects loss of these somas in eyes afflicted with retinitis pigmentosa. We achieve this by acquiring thousands of volume images, and registering and averaging them in 3D with subcellular accuracy to improve image contrast-to-noise ratio. This advance in imaging capability may provide new biomarkers for early detection of diseases that afflict photoreceptors and improve gene therapy success rates by identifying patients and retinal patches with sufficient soma populations suitable for treatment.

Pascal Rol Keynote Address: Glaucoma is a common neurodegenerative disease associated with progressive axon degeneration and cell death of retinal ganglion cells (RGCs). Here we will review recent steps forward revealing the pathophysiology of RGC degeneration, and the therapeutic targets to promote RGC neuroprotection and regeneration. Translating these advances from lab-based research in animal models, into human clinical trials with patients, has begun in earnest, but will likely rely on development of additional biomarkers for measuring relevant structural and functional outcomes in disease and its treatment. We will discuss recent data from the lab on advances in measuring RGC function in animal models in vivo, and in non-invasively measuring RGC structure, function, and metabolism in human patients.

Poor fitting contact lenses are a common cause of eye irritation, potentially leading to keratitis. We seek to improve custom contact lens manufacturing through integrating visible-light optical coherence tomography (vis-OCT) imaging and 3D printing with micrometer resolution and nanometer surface smoothness. We show that our customized contact lens fits well on various different corneas and improves the quality of OCT imaging of rodent retina.

Corneal cross-linking (CXL) stabilizes and reshapes the cornea, potentially correcting refractive errors like hyperopia and myopia. This study examines central versus peripheral CXL in six porcine eyes. Central CXL focused UV light on the central cornea, while peripheral CXL targeted the corneal periphery. Measurements using a Shack-Hartmann wavefront sensor indicated that central CXL decreased Zernike Coefficient 4, correcting hyperopia, while peripheral CXL increased it, correcting myopia. These findings suggest that selective CXL patterns can target specific refractive errors, demonstrating the technique’s versatility for precise vision correction.

Laser treatments of the retina lack objective dosing and dosing control, especially for non-damaging thermally stimulating irradiations without any visible endpoint. Over the last years, a real-time optoacoustic temperature determination has been developed and has been successfully demonstrated to determine the temperature course during retinal photocoagulation in clinical study on diabetic macula oedema (DME). In recent work, the technique has been extended towards an automatic closed loop temperature control, which is now been used in a first clinical study on patients suffering from central serous chorioretinopathy (CSCR). So far 8 patients were treated with an aim temperature of 51°C, all patients will be observed over a period of 6 months. No adverse effects owing to the treatment has been observed. As interim results, temperature control could be achieved in most laser spots and all patients showed a complete or partially resolved subretinal fluid and an increased or stable visual acuity.

Controlling tissue temperature during retinal laser therapy is crucial for predictable outcomes at non-damaging settings. We demonstrated a method for determining the temperature rise in the retina using phase-resolved optical coherence tomography in-vivo. By measuring the thermally induced optical pathlength changes during a 10-ms laser pulse, temperature rise can be detected with a precision of less than 0.5°C, enabling accurate calibration of laser power for non-damaging thermal therapy. Findings reveal significant differences in retinal deformation confinement between normal and degenerate retinas, suggesting a structural component within healthy photoreceptors that dampens the laser-induced tissue expansion, potentially serving for biomechanical diagnostics.

A subretinal photovoltaic implant (PRIMA) provided prosthetic central vision in patients impaired by atrophic age-related macular degeneration, with visual acuity of 20/420, matching its 100µm pixel size. We demonstrate, in a rodent model of retinal degeneration, that removal of such an implant and its replacement with a next-generation device having 20µm pixels increased the grating acuity to 28µm bar width, reaching their natural acuity limit. Anatomical studies demonstrated preserved inner retina after the implant replacement. These results support the feasibility of upgrading the subretinal implant in-situ to improve prosthetic vision, potentially with acuity exceeding 20/100.

Previous studies suggest early histopathological signs of keratoconus manifest in the epithelium and Bowman’s microlayer across the entire cornea. High-resolution OCT imaging may enable the detection of microstructural changes in these layers in early keratoconus. Our team developed an OCT system with a near-normal incidence beam at the cornea for wide-field imaging of the anterior layers of the cornea with high resolution. In this study, we developed a segmentation algorithm specifically designed to segment the microlayers of the cornea for the images acquired by the OCT system. This study aimed to demonstrate the feasibility of mapping the thickness of the anterior corneal layers (epithelium and Bowman’s layer) over a wide field using these OCT images.

We present an enhanced Gouy-phase Optical Transmission Tomography/ Microscopy (OTT/M) system integrated with an intelligent 3D auto-alignment mechanism for anterior eye imaging. This novel system utilizes a dual-camera setup and neural network processing for precise eye localization, replacing manual alignment. Additionally, our custom-built motorized stages significantly reduce costs while maintaining high imaging quality. Clinical applications demonstrate the system's potential to provide robust, high-resolution imaging of anterior eye structures, making it a valuable tool in ophthalmic diagnostics.

This study introduces a wide-field OCT imaging approach for the anterior corneal layers using a custom-designed hypercentric lens. The lens maintains near-perpendicularity of the probing beam to the anterior corneal surface, optimizing image contrast and resolution across the corneal diameter. The OCT system operates in dual-path and common-path modes, each with distinct advantages and limitations. Our findings show the system's ability to image and measure the thickness of anterior corneal microlayers with high resolution from the corneal center to its periphery. This approach shows promise for early detection of corneal conditions, such as keratoconus.

Structural features of the human vitreous body were assessed in vivo using high-resolution optical imaging with novel synchronous adaptive focus and OCT imaging window z-scans. Stitching together B-scans from different depths of the eye generated OCT tomograms of the whole eye. High resolution provided insight into the microstructure of the vitreous body with details that cannot be resolved by ultrasound imaging. In 84 eyes of 42 subjects aged 23-72 years, results showed age-related structural changes whose central vitreous signal intensity correlated with age (R=.612, p<0.001). This novel OCT approach enables comprehensive analysis of vitreous degeneration, enabling enhanced diagnostics in vitreo-retinal diseases.

We present the first in vivo measurements of the age-dependence of lens elasticity acquired using a system combining Brillouin microscopy and Optical Coherence Elastography. Data on 22 subjects from 21 to 69 years show that there is an overall increase in lens stiffness with age that is due to an enlargement of the central stiffer plateau observed in the Brillouin elasticity profiles. Increased lens stiffness is associated with a reduction in accommodative response. These findings suggest that age-related changes in the mechanical stiffness profile within the lens play a critical role in presbyopia.

Retinal capillary perfusion is heterogeneous and has significant spatial and temporal variations that correlate with local metabolic activity. To investigate the spatial and temporal heterogeneity of retinal capillary perfusion, we developed a novel post-acquisition image processing algorithm involving multi-volume registration followed by 3D coefficient of variation (CoV) computation to quantify perfusion variability. This method provides a more reliable representation of perfusion heterogeneity in macular circulation in a cohort of control subjects and will be applied to different patient groups to assess the potential of perfusion heterogeneity as a diagnostic indicator.

We present a multi-color scanning laser ophthalmoscope based on a supercontinuum source integrated into a Spectralis SLO. The system is capable of imaging the retina at two wavelengths simultaneously in order to extract the blood oxygen saturation values within blood vessels from the two images. To validate that our mcSLO system accurately retrieves oxygen saturation from the multi-color measurements, we have built a model eye with artificial retina and blood vessel embedded in an ex-corpus blood flow system. We have performed first measurements on the retina of a healthy subject, and will present a first analysis to extract oxygen saturation.

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. 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. Our Doppler OCT procedure provides an absolute velocity measurement which is used to properly calibrate in vivo large-area Doppler flowmetry velocity maps.

A prototype Doppler OCT device capable of fully-automatic measurement was employed to measure the structure and flow dynamics of a retinal vein at an arteriovenous crossing location. The scan location was set longitudinal to the vein and captured for 2 seconds in healthy subjects. From the OCT image, it was observed that the vein was pushed toward the outer retina by the artery. Meanwhile, it is observed that the flow velocity profile inside the vein close to the crossing point was deviated from the parabolic profile. The waveform was slightly skewed and such parameter was quantified.

The anisotropic biomechanical properties of retinal tissue may cause vessel changes in different orientations, leading to anisotropic vasodilation, as demonstrated in mouse retinas. This study verifies anisotropic vessel pulsatility in human retinas using Doppler OCT. We investigated healthy subjects, measuring vessel diameter changes in both axial and lateral directions. Circular scans around the optic nerve head revealed more pronounced pulsations axially compared to laterally in both arteries and veins. Maximum constrictions reached 34% axially versus 25% laterally in arteries, and 28% axially versus 13% laterally in veins, highlighting the anisotropic nature of retinal vascular dynamics in humans.

The choriocapillaris functions as the primary source of blood supply for the outer retina. Consequently, any pathological alterations in the choroid can affect human vision. Considering the limited comprehension of many chorio-retinal diseases, the evaluation of choroidal oxygen saturation (sO2) can assist in analyzing macular metabolism and provide insights into pathophysiology. In this study, we examined the interferogram from the choroidal region of both healthy and pathological subjects using visible light optical coherence tomography (VIS-OCT). A quantitative index reflecting sO2 was derived from the spectral hemoglobin attenuation curve, demonstrating the potential of utilizing non-invasive sO2 data in a clinical setting.

Ultrawide-field optical coherence tomography angiography (UWF-OCTA) is critical for detecting peripheral retinal vascular abnormalities in conditions like retinopathy of prematurity (ROP) and diabetic retinopathy (DR). An optimized scanning pattern can balance UWF-OCTA sensitivity to capillaries with slow flow and resistance to motion artifacts and noise. To generate high-quality 140° OCTA on a system without eye-tracking, we adopted a novel bidirectional interleave scanning pattern, which can eliminate flyback time. We also investigated two B-scan intervals (8 ms and 4 ms), in order to determine the most reasonable option for ROP imaging. Our preliminary results indicate that the UWF-OCTA with 4 ms interval can provide better performance in resolving choroidal vasculature and delivering overall clean angiograms with the least motion effects.

This study introduces high dynamic range (HDR) choroidal imaging with transcranial illumination to enable totally noninvasive choroidal imaging. The proposed method uses a near-infrared LED and oblique back-illumination, eliminating the need for contrast agents and pharmacological pupil dilation. Sequential fundus images with different illumination levels are combined to produce a high-resolution, uniformly bright image. Comparative analysis with traditional ICGA and OCT demonstrates that this technique effectively visualizes the choroid, providing image quality comparable to clinical ICGA. This non-invasive, cost-effective approach is particularly beneficial for patients in rural and underserved areas, ensuring high-quality choroidal imaging.

Introduction to the 35th Anniversary of Ophthalmic Technologies session by the conference chairs.

The first SPIE Ophthalmic Technologies conference was held in Los Angeles in 1991. It covered exciting emerging technologies including excimer laser beam delivery systems for corneal refractive surgery, laser Doppler interferometry to measure axial eye length and retinal thickness and the concept of electrode array for artificial vision. Since then, each year, the conference has continued to showcase advances and new emerging ophthalmic technologies that had a transformative impact on clinical ophthalmology. We will provide a perspective on the technologies presented at the conference, how they have evolved, their impact on ophthalmic care and the future of ophthalmic technologies.

The use of adaptive optics (AO) is driving a paradigm shift in how ophthalmoscopy is used to investigate structure and function in healthy and diseased eyes. This is because AO corrects for blur caused by optical imperfections in the eye and enables resolution on a microscopic scale, making it possible to image and interrogate the function of individual cells in the eye. It is important to note that AO is just one of many components in any given instrument. Over the nearly three decades since its first demonstration AO technology has advanced tremendously, but the advances have primarily been due to the creativity in how it has been used. AO has proven effective for multiple modalities from imaging platforms like conventional fundus camera imaging, scanning light ophthalmoscopy (including phase contrast and fluorescence) and optical coherence tomography, to vision testing platforms like AO microperimeters and AO vision simulators. In this presentation, I will highlight specific examples of how the use of AO has yielded new knowledge about retinal structure and function in healthy and diseased eyes. I will close with a look toward the role of AO in the future.

Retinal degenerative diseases lead to blindness due to loss of photoreceptors, while neurons in the inner retinal layers are largely preserved, albeit with some rewiring, which provides an opportunity to reintroduce visual information by direct stimulation of the second- and third-order retinal neurons. Several systems have been developed for electrical, optogenetic and ultrasonic stimulation of the retina and of visual cortex, and some of them have been tested clinically. I will review the current state of the field, including the recent clinical studies of a subretinal photovoltaic implant, where patients impaired by age-related macular degeneration regained their ability to read with acuity matching the 100 um pixel size: 20/400 without electronic zoom and up to 20/63 with zoom. I will also discuss the next-generation implant with 20 um pixels, which reached the natural resolution limit in rats (28 um) and may improve the prosthetic visual acuity in human patients up to 5-fold.

Optical coherence tomography (OCT) is a technology invented in 1991 to image small critical tissue structures with micrometer resolution. It is widely used in eye and coronary heart diseases. I will tell the story of OCT’s initial conception from the inventor’s perspective along with an account of the rapid pace of development that made OCT a clinical reality. The biggest applications of OCT in the management of eye diseases will be shown. Recent advances that enable OCT to advance beyond the imaging of tissue structure to the detection of blood flow and photoreceptor function will be described.

Fifty years ago, the demands for an instrument to change the shape of the cornea were clear: micrometer-precision, no touch-technique, no collateral damages like mutagenic or thermal. With the advent of the excimer laser, we learned 1981 from the micro-chip industry that at least the first 2 preconditions were fulfilled with this laser type and other groups and we (in Berlin and Dresden) did our best to find treatment parameters that minimized the thermal and actinic side-effects before we went on to clinical studies. Also, alternative wavelengths in the infrared were investigated. Once we were sure to avoid pathologic side-effects we started with purely therapeutic applications (PTK) and started in 1987 with elective interventions (PRK). In 1989, the first prospective clinical trial on excimer laser-correction of myopia was published. The first wavefront-guided treatments were conducted in Dresden in 1997.

The corneal nerves sense the state of the ocular surface, regulate ocular surface homeostasis and change in form and function in disease states. Corneal nerve dysfunction has been linked to refractive surgeries, dry eye disease, and other diseases. Understanding how the function of the corneal nerves changes in these circumstances is important both to understanding disease progression and assessing therapies. Here, we build on our previous work and demonstrate in vivo volumetric calcium imaging in the murine cornea over a large field of view using a custom spinning-disk confocal imaging system.

OCT imaging studies in rodents usually rely on the use of anaesthesia to immobilise the animals. Using a polarisation sensitive optical coherence tomography system, we investigate the effects of isoflurane and MMFK on retinal tissue dynamics using an elastographic OCT analysis. Our preliminary data of 7 mice indicate a drop of the heart rate using MMFK and an increase in heart rate using isoflurane. We also found that the pulsatile tissue displacements tend to increase with an increasing heart rate. The pulsatility is therefore generally higher for Isoflurane compared to MMFK. The strong dependence of retinal motion patterns on anaesthesia indicates the importance of considering the impact of anaesthesia.

This study presents in vivo measurements of the temporal transfer function of human photoreceptors using flicker optoretinography (f-ORG) integrated with Spatio-Temporal Optical Coherence Tomography (STOC-T). By adjusting stimulation wavelengths, the study examines variations in photoreceptor responses based on cone type. The STOC-T system enhances imaging by suppressing crosstalk noise and extending range. Measurements, taken from a healthy 28-year-old female, reveal two exponential components in the transfer function, dominant at low and high frequencies. These findings highlight f-ORG's potential in ophthalmic diagnostics.

Optical coherence tomography angiography (OCTA) provides high-contrast imaging of retinal vasculature but mainly offers anatomical rather than functional imaging. Functional imaging is essential for assessing tissue functionality, yet visualizing and quantifying light-evoked hemodynamic response in the human retina with OCTA is challenging due to motion artifacts like eye movement, blinking, and cardiac pulsatility. This work introduces a functional OCT prototype using off-the-shelf optics and a new OCTA processing method to visualize and quantify retinal hemodynamic response. Our approach statistically mitigates motion artifacts and noise, yielding more accurate results. Cross-sectional OCTA images show increased perfusion in response to light stimuli. Data from six healthy subjects indicate a consistent flow index increase (21.81±19.99%) during green light stimulation. This technique holds promise for early functional assessment of retinal circulatory impairments in degenerative diseases.

Optoretinography (ORG) is an emerging technique that records optical path length (OPL) changes at the nanometer scale in the retina in response to light stimuli. However, real-time scanning of the same patch of the retina is extremely challenging due to incessant eye movement, even during fixation. Here, we overcome eye movement limitations and demonstrate potential for single-cell imaging and single-spot optoretinography sensing (SCISSORS) in living human subjects. Using real-time eye motion correction via adaptive optics scanning light ophthalmoscopy (AOSLO), we were able to acquire ORGs in different scan modes with a level of flexibility and spatial control previously unattainable over an entire 5-second imaging session. The presented single-cell optoretinography, or SCISSORS, potentially opens new doors for vision science and ophthalmology.

The development of optoretinography (ORG) has greatly benefited from advancements in optical coherence tomography (OCT). Aiming for clinical translation, we developed a flicker-ORG modality implemented on a raster-scan OCT system that offers 13.5-µm lateral resolution and 600-kHz A-scan rate. Despite the limited scanning speed, we successfully implemented a repeated volume scan protocol, enabling examination over a finite retinal area. Additionally, the use of flicker stimulus eliminates the need for dark adaptation, thereby enhancing examination efficiency. Compared to other research-grade OCT systems, our system more closely aligns with clinical OCT specifications, offering superior cost-effectiveness and compatibility with existing clinical practices.

Optoretinography (ORG) is a non-invasive, OCT-based technique used to measure the functional response of photoreceptors to external light stimuli. Recently, a novel, velocity-based method was developed to measure these responses using low-cost, clinical-grade OCT components, with the goals of facilitating adoption of the technology and improving patient throughput. Here, we show that the velocity-based approach can simultaneously measure responses to stimuli within the photoreceptor outer segment and in the extracellular space just distal to it. The latter measurements are of interest for two reasons. First, they may shed light on proposed mechanisms of the outer segment response, which contain untested assumptions about water movement into and out of the outer segment. Second, they may have independent clinical value as indicators of fluid movement, which is thought to be impeded in some degenerative diseases of the outer retina.

This paper introduces a method and device for ensuring correct eye positioning during ophthalmic examinations. The system enhances accuracy and efficiency by monitoring the eye's position in real-time using a CMOS camera and a Support Vector Machine (SVM) algorithm. This algorithm classifies the eye's position and triggers a visual or auditory warning when optimal conditions for image acquisition are met. Applicable to various ophthalmic tests, including Optical Coherence Tomography (OCT), visual field analysis, corneal topography, and pupillometry, the method involves collecting eye images through wavelength-specific infrared illumination. A dichroic mirror separates the acquisition from the original system used for stimulation. The SVM classifies images while analyzing gaze, blink, and eye position. Preliminary tests optimize the classification algorithm for embedded computers, achieving millisecond processing times on a 10-watts ARM CPU. The system can be integrated into new ophthalmic devices or retrofitted into existing ones, aiding researchers and medical experts in performing accurate and reliable eye examinations, especially with non-collaborative patients.

This study presents computationally enhanced retinal optical coherence tomography (OCT) imaging to address challenges in high-resolution retinal imaging. The research focuses on two main aspects: computational aberration correction (CAC) and computational de-speckling (CdS). CAC is developed based on OCT image formation theory to overcome depth-dependent aberrations in volumetric imaging. CdS, enabled by CAC, is founded on speckle formation theory to improve image quality without requiring multiple acquisitions or sophisticated motion correction algorithms. The combination of these techniques, demonstrated using a 1-μm swept-source OCT, results in high-quality, fine retinal structural imaging. This approach offers a low-cost method for assessing microstructural retinal biomarkers, potentially advancing retinal diagnostics and research.

A novel diffusion-based solution is developed for resolution enhancement and noise removal in optical coherence tomography (OCT) images. The proposed method conditions a diffusion model (DM) using physical priors of an OCT system to inverse the total image degradations due to the speckle noise, defocus blur, digital sampling, and light diffraction. The performance of the proposed method was tested using in vivo cellular-resolution images of human cornea acquired with a line-field OCT (LF-OCT) system, resulting in enhanced visibility of cell boundaries and corneal nerves. The developed solution relies solely on the OCT intensity data, offering the potential to provide ophthalmologists with a viable tool for super-resolution imaging of the eye.

Retina experiences decreased ganglion cell (GC) density and reduced dendritic and axonal areas by its aging. Glaucoma also leads to GC loss and irreversible blindness. We developed a deep-learning-based scatterer density estimator (SDE) to assess the aging of retinal tissues. Analyzing OCT data from 134 eyes (78 healthy, 56 glaucoma), we found significant age-related reductions in ScD across all retinal layers. ANCOVA revealed no significant differences in ScD between normal and glaucoma groups when adjusting for age, but retinal thickness showed significant differences. The SDE effectively measures age-related changes, with ScD indicating retinal aging and GC thickness distinguishing glaucoma.

Subretinal delivery of novel gene and cell therapies currently relies on either visual estimation of bleb size, injection duration, or microinjector readings to determine injected volumes, all of which result in high variability of drug administered. Intraoperative optical coherence tomography (iOCT) can achieve 4D imaging of surgical dynamics at video-rates when combined with high-speed sources and optimized scanning waveforms and shows promise in providing more accurate and reproducible volume quantification metrics. In this work, we demonstrate 4D imaging of bleb formation in retinal phantoms and present segmentation algorithms to quantify bleb volumes during simulated subretinal injections.

This study aims to develop and validate a cost-effective, non-mydriatic smartphone pediatric camera (PedCam) for comprehensive eye examinations of newborns and young children. Diagnosing pediatric retinal diseases like retinopathy of prematurity (ROP) requires a thorough examination of the retina's periphery. The PedCam captures ultra-widefield images with a 180° FOV in a single snapshot using a portable transilluminator and off-the-shelf lenses. A detachable, 3D-printed housing facilitates easy smartphone attachment, while image acquisition software aids in automatic focusing and image capture. Zemax simulation and experimental validation confirm the imaging performance. Comparative imaging with a clinical ultra-widefield SLO demonstrates the PedCam's superior capability. This innovation offers a portable, efficient, affordable solution, enhancing access to essential retinal care for remote and underserved populations.

The slit lamp is used for examination of the ocular anterior segment by projecting a homogenous, adjustable slit of light onto the eye. Conventional slit lamp systems require specially trained operators and patients to be stabilized, therefore limiting access to certain groups of patients. Expanding on previously developed robotic optical coherence tomography techniques, we created a robot mounted slit lamp with color cameras for data capture. We then further optimized the optical system for dynamic illumination. Here, we explain system design, demonstrate in vivo capture of the robotic slit lamp, and describe dynamic illumination during pupil tracking.

Optical coherence tomography is an indispensable imaging modality for the diagnosis and management of eye disease but is deployed almost universally as a tabletop device in ophthalmology specialty clinics that requires patient who can sit upright and participate in imaging. We introduce a mobile robotic OCT prototype designed for routine imaging in outpatient and inpatient settings and characterize its suitability for imaging in diverse clinical configurations. We validated the system's workspace and imaging configurations for standing, seated, and reclined postures and demonstrated the effectiveness of predictive cancellation using a retinal eye phantom on a motorized stage to simulate jerk nystagmus.

Cataracts are the leading cause of blindness worldwide, and cataract surgery is the most commonly performed surgery in the United States. Ophthalmic surgeries such as cataract surgery involve precise 3D surgical maneuvers that can benefit from the depth visualization capability of OCT. Unfortunately, current surgical OCT systems are hampered by extensive manual lateral and axial tracking of the OCT scan volume to the surgical instrument. We have developed a flexible tracking software platform for surgical OCT systems that uses the stereo microscope video feed to perform automated lateral and depth tracking of OCT to the surgical instrument tip.

Large Language Models (LLMs) enhance the safety and efficacy of AI-assisted diagnosis and treatment through interactive human-computer dialogue. However, due to the imperative to protect patient privacy and ensure data security, OCT devices in healthcare settings are commonly operated offline. Additionally, the operation of current LLMs demands substantial graphics memory. In this paper, we introduce ChatOCT, an embedded clinical decision support system with specialized knowledge in OCT, designed to operate offline and with minimal computational resources. The superiority of our ChatOCT, has been independently confirmed by ChatGPT (GPT-4) and two clinical ophthalmologists.

In this approach, visual patterns on a display are used to animate the patient to look at specific fixation points. The aim is to guide the patient's view through a followable target and direct it over the region of interest. The resulting fundus images are processed and tracked. In contrast to conventional systems, which require static fixation, our system uses the natural eye movements. The system is designed to enable precise mapping and acquisition of OCT A-scans based on the fundus images in order to produce accurate optic nerve B-scans. The entire setup is primarily based on commercially available components, with the aim of developing a cost-effective alternative to traditional devices.

Anterior segment ischemia is a rare but potentially vision threatening complication to strabismus surgery that arises due to damage to the anterior ciliary arteries that run along the rectus muscles. To reduce the risk, a maximum of two rectus muscles are operated on at a time. Previously, no method has been able to provide real time date on perfusion during strabismus surgery. We have demonstrated that Laser Speckle Contrast Imaging can be used to visualize perfusion in real time, non invasively during strabismus surgery. We have measured anterior segment perfusion in 96 strabismus surgeries before- and after rectus muscle detachment. Data has also been collected from 24 enucleations to enable perfusion monitoring as all four rectus muscles are detached. Perfusion decreased significantly following surgery on both horizontal- and vertical rectus muscle detachment but was more pronounced in the latter. In the enucleations, perfusion remained stable until the fourth rectus muscle was detached, indicating that surgery on three rectus muscles may be feasible. LSCI may thus be a valuable tool in reducing the risk of anterior segment ischemia following strabismus surgery.

A quantitative analysis technique for the biomarker of dry eyes was developed to enable objective diagnosis and personalized treatment. We developed the fluorescence imaging system with fluorescein sodium and 480nm LED light source to monitor the tear film change and corneal epithelial erosions (CEE). Fluorescence images need to be segmented to selectively analyze the fluorescence signal in the cornea. The U-net model was trained because the shape of the anterior segment of the eye varies with patient and time. Changes in the tear film breakup were extracted from the segmented images and defined by the area ratio. The corneal diameter was measured as the maximum length from the largest contour of the segmented images, and the NEI grading scale was applied to extract corneal staining score (score of CEE) by region.

Optical Coherence Tomography (OCT) is a non-invasive, high-resolution imaging technology extensively used in ophthalmology to diagnose and monitor retinal diseases such as diabetic retinopathy, macular degeneration, and glaucoma but it lacks the capability to visualize blood flow, which led to the development of OCT Angiography (OCTA). OCTA provides clear visualization of the retinal capillary networks but requires separate hardware, takes longer time to scan and is affected by motion artifacts. Hence it has not been integrated into widespread clinical workflow and generating OCTA images from OCT through generative machine learning can be a promising alternative. The purpose of this study was to demonstrate a quantitative comparison of vascular features on Optical Coherence Tomography Angiography (OCTA) images translated from OCT vs. commercially available OCTA.

Retinal degenerative diseases are leading causes of incurable blindness. Although photoreceptors are impaired in these diseases, inner retinal neurons retain some functionality, thereby providing an opportunity to simulate a vision-like experience by activating the remaining neurons. Ultrasound is an especially intriguing potential stimulus since it has been reported to have neuromodulatory capability without the need for invasive approaches. In this study, polymer matrix nanocomposites (PMNCs) were implanted in front of the retina in rabbits and activated with a 1064 nm laser system to generate acoustic waves, aiming to stimulate retinal neurons. Our result showed that the implanted PMNCs were situated at a distance of approximately 200µm from the retinal surface in OCT images. The visual evoked potential (VEP) showed low-amplitude atypical waveforms following photoacoustic stimulation. These findings suggest that PMNCs has the potential to serve as a novel photoacoustic transmitters for epiretinal prostheses.

Reactive oxygen species (ROS) are implicated in the pathogenesis of common blinding diseases. We utilized a novel autofluorescence multispectral imaging (AFMI) technique for the non-invasive diagnosis of ROS in a human retinal pigment epithelium cell culture model. The AFMI achieved high ROC AUCs above 0.85, distinguishing healthy control cells from those exposed to various metabolic stressors. This study highlighted AFMI's potential as a clinical imaging tool for diagnosing neurodegenerative diseases like glaucoma and age-related macular degeneration.

This research aims to develop an Auto-PIMD program for calculating Pigment Epithelium – Inner limit of the retina Minimal Distance Averaged Over 2π Radians (PIMD-2π) lengths in OCT images to aid in glaucoma diagnosis. Utilizing the nnU-Net model for semantic segmentation, the study focuses on improving Optic nerve head Pigment epithelium Central Limit (OPCL) calculation accuracy. Experiments using 78 OCT images from Uppsala University compared nnU-Net cylindrical and Cartesian architectures against the 2D U-Net framework. Results show the nnU-Net frameworks produce more accurate OPCL coordinates and eliminate manual selection biases. All methods effectively distinguish glaucoma patients from healthy subjects, with nnU-Net demonstrating greater stability. This study highlights the potential of the Auto-PIMD program in clinical applications.

There is currently no non-invasive, low-cost, and scalable solution for tracking molecular concentrations of important early-warning proteins, such as vascular endothelial growth factor (VEGF) and tumor necrosis factor alpha (TNF-alpha) within the eye. Our previous work showed the ability to use a Fabry-Perot cavity on a keratoprosthesis to stably and non-invasively monitor intraocular pressure (IOP) for over 4.5 years. Here, we present a bifunctional flexible metasurface array that combines aptamer-based sensing sections with the aforementioned IOP monitor, which is aimed for integration onto an intraocular lens. We tested the simultaneous pressure and biochemical measurement capability in pressurized artificial aqueous humor against gold-standard ELISA assay measurements. The results will contribute to a process for designing a multiplexed longitudinal biochemical monitoring platform, creating insight into how relative concentrations fluctuation can help create early diagnostic tools.

Blood flow measurements are important in many applications, especially in the retina. In the clinic, flow is typically imaged with optical coherence tomography angiography (OCTA), fluorescein angiography or laser speckle contrast imaging. While conventional non-invasive imaging modalities can visualize flow, they cannot determine the absolute flow without information about the angle between the flow and the probing beam. Typically, two or three beams are required in OCTA to determine the absolute flow. Measurements on the flow phantom demonstrated our speckle imaging system can quantify the absolute flow independent of the angle between the vessel and the imaging system. Measurements in the human eye demonstrate the method can measure the flow in the upper layers of the retina, and in choroidal blood vessels below the photoreceptors-retinal pigment epithelium-Bruch's membrane complex.

Early identification of ophthalmic disease can lead to timely interventions, potentially preventing or minimizing vision loss and improving patient outcomes. Proper classification of ophthalmic disease ensures that patients receive accurate diagnoses which is essential for effective treatment planning. This study introduces an innovative deep learning (DL) framework that uses recent advancements in diffusion models to classify various pathological retinal disease datasets. The performance is compared to traditional baseline methods using optical coherence tomography (OCT) images. Our framework comprises two networks: one for training a model for the reverse diffusion process and the other for semantic encoding to condition the denoising process and improve the outcomes of the first model. With these two networks, we pre-trained a semantic encoder that captures the most prominent features of OCT images and then we compared the pre-trained semantic encoder with a baseline model on six different multi-class classification tasks. Our findings show the superiority of the pre-trained model and indicate the possibility of using the diffusion process as an unsupervised pre-training phase.

Choroidal neovascularization (CNV) is a leading cause of vision impairment in wet macular degeneration. Current treatments like intravitreal anti-vascular endothelial growth factor (VEGF) therapy with drugs such as bevacizumab (BEV) require frequent administration and carry risks like endophthalmitis. This study explores a novel approach using ultrsmall biodegradable silicon nanoneedles (SiNNs) integrated into a tear-soluble contact lens for treating CNV. SiNNs encapsulated with BEV (BEV@SiNNs) were developed for sustained drug delivery. In a rabbit CNV model (n = 7), CNV was induced by subretinal injection of Matrigel and VEGF. A subconjunctival contact lens with SiNNs was placed on the posterior sclera 3 days post-CNV induction. The treatment's efficacy was monitored using color fundus photography, photoacoustic microscopy, OCT, and fluorescein angiography (FA) before treatment and at specified intervals up to 12 months. SiNNs demonstrate potential as an effective platform for long-term, sustained treatment of CNV in this rabbit model, highlighting their suitability for extended drug delivery applications.

Continuing our previous work on linear optical coherence tomography (LOCT) for ophthalmic applications, we present a method to extend the measurement range to address the limitation of low measurement depth. By optimizing the configuration of the reference arm, we achieved a significant improvement in measurement depth while maintaining high image resolution. The performance of the updated LOCT system was validated through comprehensive testing on an artificial eye model, demonstrating an increase in measurement depth compared to the former system. In addition, a visual comparison was made between a B-scan of the LOCT and a conventional state-of-the-art spectral domain OCT to make a rough classification of the quality of LOCT.

The objective of this work is to display that fluorescein signal attenuation in retinal arteries compared to retinal tissue creates retinal blood flow (F) overestimation. F is extracted from human fluorescein videoangiographies using dynamic tracer kinetic modeling (DTKM); DTKM-extracted F has been estimated at values 10-20 times higher than previous human studies. Using Beer Lambert Law and optical properties at 570 nm, the ratio of retinal tissue and blood’s detected fluorescein intensities was found to be between ~12-17. The retinal artery’s diameter had little effect on the ratio of retinal tissue and blood’s detected fluorescein intensities, however, the retinal tissue thickness was proportional to the increases in this ratio.

Retinal imaging systems must be regulated and evaluated using specialized and standardized phantoms to ensure precise and consistent patient imaging. With its exceptional capability to produce phantoms of nearly any complexity, 3D printing is a powerful tool for rapidly prototyping phantoms. However, fully 3D-printing eyeballs remains challenging due to the complexities involved in producing optical components and achieving precise alignment. In our current study, by using a robotically controlled multi-material 3D printing system, we can fabricate eye phantoms with desired shapes and optical profiles, and the eye model can be used as a testing target for quality control and functional verification of the developed ophthalmic imaging systems.

We propose a method for dynamically adjusting the reference arm position based on the shape of the retina, enabling deeper depth range imaging with a wide field of view. Our technique utilizes a novel scanning method, where a small 4.5 mm diameter circular B-scan moves spirally from the center to the retina’s periphery, covering an 80-degree field of view. This approach is notable for its ability to make image flipping unlikely occur by adjusting the reference arm position, as demonstrated in eyes inlcuding those with high myopia. We also present an image reconstruction method to compensate for eye motion and reference arm adjustments.

We present a software-based approach for real-time feature tracking of funduscopy images. Our method includes an initial preprocessing step that enhances contrast, accentuates retinal details, and removes reflections originating from the eye and funduscopy setup. Subsequently, feature detection, description, matching and filtering are performed. These steps are optimized to ensure low error rates and robustness, even in low quality video signals. The results demonstrate real-time tracking with high detection rates and minimal misdetections. Although our primary application involves tracking the positions in a scannerless linear optical coherence tomography system, this approach is also applicable in other fields requiring fast and precise retinal movement detection.

Low-cost optical coherence tomography (OCT) lowers the financial barrier for entry in ophthalmic diagnostics at-the-point of care, especially in low-resource environments. Gen 3 low-cost OCT offers improvements to previously published generations including leveraging the reduced cost of 3D printed spectrometers to enable balanced detection. Preliminary results from a clinical trial compare Gen 3’s macula imaging quality in healthy patients, quantified via CNR, to previous generations and a commercially available system. Gen 3 average CNR was measured to be 1.924 ± 0.024 over 221 macula images, higher than previous generations of low-cost OCT and a commercial system, 1.803 ± 0.036.

We propose a theoretical wave-optics model to simultaneously investigate the influence of aberrations and spatial coherence of the illumination on the lateral resolution of a full-field optical coherence tomography (FFOCT) imaging system. In the scope of retinal imaging, this model can guide us to search for an Adaptive-Optics strategy and a degree of spatial coherence of the source that associates high resolution, high signal-to-noise ratio (SNR), while keeping the system compact and simple.

Image registration plays a pivotal role in enhancing the diagnostic accuracy of ultra-widefield (UWF) retinal images, despite challenges like morphological warping due to the microsaccades and inherent imaging inconsistencies at different imaging time points. In this work, we successfully registered UWF-OCT images by combining 3D vision techniques with conventional registration methods for motion correction and alignment. The precise alignment of images taken at multiple time points enables clear visualization of subtle retinal changes, enhancing patient follow-up and disease monitoring. The proposed method offers more accurate tracking of morphological changes, leading to improved diagnostic insights and better clinical outcomes.

At the core of retinal disease management lies high-quality fundus images, which facilitate disease diagnosis, progress monitoring, and treatment delivery. Existing imaging platforms can be expensive that may be unaffordable in developing, low-resource regions. To address this need, we explore the feasibility of using smartphone-based imaging systems to acquire high-quality fundus images at reduced cost. In particular, our study focuses on the software side, aiming to automatically output fundus images with high image qualities to ensure the entire imaging system is maintained at a lower cost. Using anesthetized Long Evans rat fundus videos, we show our model can effectively distinguish frames with various image qualities, showing promise for smartphone-based video frame selection.

This paper introduces a semi-automatic alignment module tailored for next-generation ophthalmic imaging systems, specifically Optical Coherence Tomography (OCT). The module addresses critical challenges in modern ophthalmic diagnostics, such as space efficiency, patient comfort, and precise alignment. Our design integrates head stabilization, three-dimensional optical axis matching, and real-time feedback, ensuring accurate and repeatable imaging. Experimental results validate the system’s effectiveness, demonstrating enhanced diagnostic precision and ease of use. This alignment system offers a comprehensive solution for seamlessly incorporating advanced imaging technologies into existing clinical settings without compromising patient experience.

Retinal diseases are significant causes of vision impairment, underscoring the need for early and accurate diagnostics to prevent irreversible vision loss. Optical coherence tomography (OCT) plays a critical role by offering high-resolution retinal imaging. This study introduces a 3D semantic segmentation model based on the nnUNet (no-new-UNet) architecture to accurately identify retinal features. The model aims to generalize across diverse public OCT datasets, improving diagnostic accuracy and speed. Unlike previous 2D methods with limited categories, this approach involves detailed annotation and rigorous review to ensure accuracy. Preliminary results indicate that the model performs comparably to expert diagnosticians, highlighting the importance of accessible, well-labeled OCT datasets for advancing AI-driven medical tools.

Developing an accurate optical model of the mouse eye is essential for advancing ophthalmic research. However, existing models often lack specificity, failing to capture the unique optical properties of different eyes. In response to these challenges, we developed a three-step method using an electronically tunable fundus camera to improve the accuracy of the Remtulla model, a widely recognized optical model of the eye. When applied to wild-type mice, our approach significantly enhanced the alignment between the model and actual observations. Despite these advancements, the method requires precise calibration and alignment. Future work will focus on refining these processes to achieve greater robustness and reliability.

The combined Multimodal Scanning Confocal Light Microscopy (SCLM) with Optical Coherence Microscopy (OCM) system has been developed for ex vivo mouse retina tissue imaging. This instrument utilizes a commercial microscope and our previously reported multimodal SLO-OCT acquisition engine, allowing simultaneous detection of back-scattered light from the SCLM illumination beam, fluorescent emission, and depth information about different retinal layers by means of OCM. Identical data formats of in vivo and ex vivo measurements facilitate better data analysis from the same animals. In this presentation, we will review the design details of the combined SCLM/OCM system, showcase its performance, and compare retinal morphology acquired in vivo and ex vivo from the retina of the same animal.

A versatile Optical Coherence Tomography (OCT) system has been designed to overcome the limitations of current fragmented OCT technology. The system captures high-resolution images of all ocular structures in a single device, including the retina, cornea, and lens. The system's results include a scan range of 9 mm on the retina and 4 mm on the cornea. Additionally, the system can estimate axial lengths, anterior chamber depth, and lens thickness with high repeatability, making it a practical solution for modern clinical practices. The system can also function as an optical biometer, providing accurate measurements for refractive surgery.

This study presents a novel approach for screening eye diseases using an automated Optical Coherence Tomography (OCT) system that operates without human intervention. Motivated by the need for reliable and consistent diagnostic tools, our system employs a convolutional neural network (CNN) to detect and correct artefacts in OCT images autonomously. The automated system aims to enhance diagnostic accuracy by eliminating operator-induced errors and ensuring high-quality imaging. By integrating advanced CNN techniques for artefact detection and image correction, the system strives to improve the early detection of eye diseases, thus advancing patient care and outcomes.

Introduction: Acidic corneal injuries cause significant ocular damage and vision impairment. This study evaluates potential-driven electrochemical clearing (P-ECC) for treating acidic corneal injuries in ex vivo porcine eyes. Objective: Assess P-ECC effectiveness, impact on tissue structure, injury progression, and clinical potential. Methods: Ex vivo porcine eyes were exposed to 5M HCl, followed by phosphate buffered saline (PBS) irrigation. P-ECC was applied using platinum electrodes. Efficacy was assessed through optical coherence tomography (OCT) and second harmonic generation (SHG) imaging to analyze corneal clarity, thickness, and collagen organization. Results: P-ECC restored optical clarity after HCl exposure. OCT showed injury progression and post-ECC recovery. SHG images indicated minimal collagen fibril rearrangement, reflecting improved clarity. Corneas in PBS alone showed no clarity restoration. Conclusions: P-ECC shows potential for treating acidic corneal injuries, restoring clarity without irreversible fiber damage. Further research is needed to understand the mechanism and optimize treatment.

We describe a new approach of presbyopia correction by using the specificity of natural human vision that is limited in viewing angle. This allows us to perform the required correction within that viewing angle at a time by using local dynamically tunable liquid crystal lenses. The operation principles of these lenses and experimental results will be presented. Our human tests show that almost a 2020 vision is recovered by using this approach.

Accurately interpreting Optical Coherence Tomography Image (OCT) images is crucial for assessing retinal diseases. However, manual analysis of OCT images remains a time-consuming and labor-intensive process and is prone to inter-observer variability. Here we analyze all state-of-the-art architectures based on interpretability and compatibility in resource-constrained devices. We introduce a novel interpretable model using the Kolmogorov Arnold representation theorem to make the models' decisions more interpretable. Our contributions aim to bridge the gap between deep learning-based solutions and clinical practice, and offer a more efficient, accurate, and interpretable OCT analysis framework.