We present a novel platform for eye tracking showing high speed and accuracy in a wide range of realizable visual tasks. The optical setup consists of the scanning laser ophthalmoscope, the actual tracker (the FreezEye Tracker), and a visual projector for task presentation. The MEMS-based tracker scans the retina with a framerate of 1.24 kHz, providing high angular and temporal resolution. Advanced algorithms allow for precise reconstruction of the eye trajectory covering the range of movements from small microsaccades to high amplitude saccades. The high quality of the generated data provides an abundance of data potentially useful for diagnostic purposes.

Current state-of-the-art human retinal vis-OCT systems have a limited field of view due to the large curvature of the retina and the system sensitivity roll-off. Here we presented the first linear-in-k visible range OCT spectrometer and achieved a sub-5dB roll-off over the entire imaging depth. The first full-range vis-OCT imaging was demonstrated to extend the system imaging depth. Thanks to 140 nm bandwidth in visible light range, the system axis resolution can achieve 1-2 um in air and a 65˚ by 65˚ wide-field human retinal full range vis-OCT imaging was demonstrated.

In this paper, we present the preliminary results of the scotopic luminosity curve for two-photon vision measurements in the spectral range from 872 nm to 1027 nm. The results were obtained thanks to a newly-developed custom-build tunable femtosecond erbium-doped fiber laser that pulse train parameters and spectral width are close to constant while tuning. Such instrumentation enabled us to perform reliable measurements across the laser tuning range of over 150 nm.

We enhanced the capabilities of an Adaptive Optics Flood Illumination Ophthalmoscope to achieve multimodal imaging. The control of the illumination with a digital micromirror device, combined with a wide field of view and a light data processing, allows us to improve the brightfield contrast with a pseudo-confocal mode and to visualize transparent retinal structures with phase contrast imaging. We achieved to get from a single sequence of acquisition, four wide field images, each corresponding to one type of the following imaging modes: contrast-enhanced brightfield, darkfield, offset aperture and split-detection imaging.

Multiple sclerosis (MS) is a debilitating autoimmune disease characterized by lesions found in different regions of the central nervous system caused by overactive immune cells. MS also manifests in the retina, in which optic nerve pathology (such as optic neuritis) and neurodegenerative processes can affect inner retinal cells and structures. We observed the inner retina to be profoundly affected by MS, including nerve fiber bundle thinning, enlarged and lower density retinal ganglion cells, and the presence of microcysts. Longitudinal quantification of inner retinal changes enabled by AO-OCT may help accelerate the development of novel therapies for MS patients.

Refractive errors commonly cause vision impairments; current treatments are temporary and/or involve side-effects. We performed electromechanical reshaping (EMR), a potential laser eye surgery alternative, to determine its effect on reshaping corneal curvature in ex vivo rabbit corneas. Optical coherence tomography (OCT) was performed to ensure corneal curvature change. Second harmonic generation (SHG) was used to determine if collagen damage occurred. After EMR, shape change was visualized with OCT while SHG indicated weaker collagen signals in certain regions. EMR is a potential treatment for refractive errors. However, more experiments are required to determine efficacy and potential for clinical application.

We present a deep-learning based approach for automated qualitative assessment of lesion volumes using OCT images to enable real-time assessment of injury severity and longitudinal tracking of tissues response to photodamage. The network has been trained to quantify photodamage between the outer plexiform layer (OPL) and retinal pigmented epithelium (RPE) accurately without the need for extensive image pre- and post-processing. Manually annotated OCT cross-sections were used as ground-truths to train a U-Net convolutional neural network. The network was designed and implemented in PyTorch based on the multi-scale U-Net architecture.

In this research, we demonstrated a CNN-based DOPU estimation algorithm without polarization-sensitive OCT signals. The CNN was trained with pairs of retinal OCT (input) and DOPU (teaching) images. The recall and precision of RPE abnormal appearances between true DOPU and synthesized DOPU of pathological eyes were calculated. For normal eyes, the grader evaluated the soundness of the RPE appearance for true DOPU and synthesized DOPU. The recalls are relatively good (0.74-0.95), while the precisions highly depend on the types of abnormalities (0.37-0.98). Five RPE abnormalities are found in synthesized DOPU within 25 synthesized DOPU B-scans while there is no abnormality in the true DOPU.

The pupil light reflex (PLR) for focal chromatic stimuli was tested in 125 healthy middle-aged subjects offspring of Alzheimer Disease (AD) patients and 61 age-matched controls,. Machine learning algorithms identified features associated with PLR latency with an Area Under Curve of 0.91±0.05 in the left eye and 0.88 ± 0.05 in the right eye. Parameters associated with the contraction arm of the PLR were more discriminative compared to parameters associated with the relaxation arm. This study suggests that subtle changes in pupil constriction latency may be detected decades before the onset of AD clinical symptoms using a simple, non-invasive chromatic pupilloperimetry test.

Microscope-integrated optical coherence tomography (MIOCT) is an emerging multimodal imaging technology in which live volumetric OCT (“4D-OCT”) is displayed simultaneously with standard stereo color microscopy. 4D-OCT provides ophthalmic surgeons with many visual cues not available in microscopy, but it cannot serve as a replacement due to lack of color features. In this work, we demonstrate progress toward a unified solution by fusion of data from both modalities, guided by segmented 3D features, yielding a more efficient visualization combining important cues from both modalities.

Critical flicker frequency measurements are important in determining the spatial variation of flicker sensitivity in the human retina. An objective and localized measurement of the frequency response of photoreceptors could help elucidate the true physiological mechanisms responsible for such flicker sensitivity variations. Flicker optoretinography (ORG) may be a promising technique for this purpose. In this work, we use Spatio-Temporal Optical Coherence-Tomography to capture flicker optoretinograms to visible patterned light stimulation modulated in the range from few Hz to 30Hz over 5.7º x 2.8º of the retina, at several foveal eccentricities, highlighting the prospect for objective flicker perimetry with ORG.

In several retinal pathologies, such as AMD and retinitis pigmentosa, function of cones and rods are affected to different extents. To investigate these, an imaging modality is needed that separates the function of both photoreceptor types. Recently, functional imaging of single cones using phase-sensitive OCT has been demonstrated. However, until now, functional imaging of rods was only possible with high resolution systems. Here we differentiate the functional response of rods and cones without resolving single photoreceptors. After stimulation with short illumination in the mesoscopic range, the different temporal dynamics of rods and cones are distinguished by their typical response characteristics.

The structure and composition of the corneal nerves in health and disease have been extensively studied, however, study of corneal nerve function in living systems has been challenging and limited. Here, we demonstrate non-contact, longitudinal imaging of in vivo murine corneal nerve signaling and in vivo stimulus-response. These developments have the potential to allow for new studies of changes to corneal nerve function in disease and damage, and for better assessment of therapies to treat dysfunction and diminished nerve function.

We introduce reflective mirror-based line-scan adaptive optics line-scan OCT, optimized for imaging light-induced retinal activity (optoretinography) and weak retinal reflections at the cellular scale. The performance was exemplified by cellular-scale visualization of retinal ganglion cells, macrophages, foveal cones, and rods in human observers. Light-evoked optical changes in foveal cones were observable at an eccentricity 0.3 deg. from the foveal center, enabling the first in vivo demonstration of reduced S-cone (short-wavelength cone) density in the human foveola. Given the challenges typically associated with optical accessibility in the living human fovea, this instrument holds significant promise for basic and translational applications.

Development of high resolution functional retinal imaging tools is of great interest to clinical and experimental ophthalmology because the alterations in retinal function, at cellular resolution, hold the promise of being more sensitive for disease diagnostic then the purely retinal morphology-based assays. In this work we present our initial design and implementation of mouse retinal imaging system that incorporates full-field (FF) swept- source (SS) optical coherence tomography (OCT) with dedicated light stimulation channel for high-speed measurements of light evoked responses in photoreceptors of mice. The Optoretinograpahy (ORG) results acquired with our FF-SS-OCT system are compared with those acquired with our standard raster scanning mouse ORG-OCT system.

Optical Coherence Tomography (OCT) serves as a high-resolution imaging technology in many disciplines. In medicine, its use in glaucoma has revolutionized the diagnosis of disease and the detection of disease progression. OCT provides objective, quantitative analysis of the retina and optic nerve, allowing clinicians to accurately assess glaucoma in ways not possible prior to its invention. This has aided countless patients, as it allows a non-expert observer to provide expert level appreciation of the patient’s disease state. In this way OCT has flattened the earth, expanding access to skilled evaluation of glaucoma worldwide. Now commercially available advances, such as widefield OCT and OCT angiography (OCT-A) using high speed swept-source imaging, already allow clinicians to make more complete assessments of their patients. OCT has continued to develop beyond measurement of retinal layers and vasculature, with laboratories now investigating the use of OCT for retinal single cell assessment, sublayer thickness quantification, optophysiology and retinal oximetry. The introduction of new iterations of OCT, such as adaptive optics OCT (AO-OCT) and visible light OCT (vis-OCT) have enabled the investigations above. Finally the use of artificial intelligence for OCT has promises exciting advances in OCT image and data analysis and in the prediction of disease activity.

Photoreceptors are the primary sensing element of the human visual system. Traditionally, photoreceptors are imaged with hardware-based adaptive optics (AO), which compensate for eye aberrations. However, these systems can be challenging to maintain. Here, we demonstrate the spatiotemporal optical coherence tomography (STOC-T) as the new modality for high-speed, cellular-level volumetric imaging of the human retina in vivo. The cellular features become visible after applying digital aberration corrections. We also show that STOC-T provides cross-sectional images (B-scans) and, concurrently, high-resolution wide-field en face images of the inner and outer human retina layers.

Sight can be restored in patients who lost photoreceptors due to atrophic AMD by substituting them with a photovoltaic array. Subretinal pixels convert pulsed NIR light projected from augmented-reality glasses into electric current, stimulating the nearby inner retinal neurons. Patients with such implants can simultaneously use their residual peripheral sight and central prosthetic vision, and its acuity closely matches the 100um pixel size of the implant. We present two approaches to selective neural stimulation with pixel sizes down to 20um – optically configurable current steering and 3-dimensional honeycomb-shaped electrodes, both providing prosthetic acuity matching the natural resolution in rats.

Full-field swept-source optical coherence tomography (FF-SS-OCT) and laser Doppler holography (LDH) are two holographic imaging techniques presenting unique capabilities for ophthalmic applications. We report on adapting an FF-SS-OCT instrument with minimal effort to allow both OCT and LDH imaging at a fast alternation rate. We used this instrument to monitor retinal blood flow with both modalities to better understand the physiological signal measured with each technique. The combination of these two imaging techniques holds potential in ophthalmology for basic research and clinical diagnosis.

A directional optical coherence tomography (OCT) was implemented to access both amplitude and sensitivity (directionality) of directional backscattering of the retinal pigment epithelium (RPE) cells. Intensity and directionality of RPE scattering from different mouse strains with varied melanin pigmentation levels (heavily pigmented, moderately pigmented agouti, and melanin-lacking albino mice) are measured and compared. Amplitude of RPE scattering is directly proportional to melanosome concentration, whereas the directionality is inversely related to the concentration of RPE melanosome. The diagnostic capability of this approach was illustrated using a mouse model of Stargardt disease (Abca4-/-), where a reduced melanin-pigmentation relative to wild-type control was detected..

Despite the rapid development of OCT, high-resolution in vivo imaging of human eye with penetration into deeper retinal layers and choroid is still a major challenge due to its sensitivity to coherent noise, such as speckle and crosstalk. To address that, we have developed a technique termed Spatio-Temporal Optical Coherence Tomography (STOC-T) that uses light with controlled spatial and temporal coherence to obtain high-contrasted coronal projection images of the choroid at various depths including that of choriocapillaris. It can also detect blood flow and reveal vascular networks in various chorioretinal layers that are otherwise invisible to OCT.

Accurate determination of the attenuation coefficients of retinal structures would allow for quantitative tissue characterization and can be calculated from OCT data. This requires compensating for the system’s confocal function. We present measurement series for extraction of the focal plane and apparent Rayleigh length from the ratios of OCT images acquired through a range of different focus depths. The optimal combination of focus depths is determined for intralipid and titanium oxide phantoms with different scatterer concentrations and the confocal-function-corrected attenuation coefficients are calculated. We further demonstrate good reproducibility in a multi-layered titanium oxide phantom and apply this method to in-vivo retinal data.

Although optical coherence tomography has explored coherent volumetric contrast extensively, the access of volumetric incoherent contrast in the human eye has been challenging. To exploit incoherent contrast in human retina in 3D, a confocal oblique scanning laser ophthalmoscopy (CoSLO) is proposed. It is the first time that the incoherent cross-sectional images of the human retina are obtained directly through optical imaging. By utilizing scanned light sheet, depth signals are acquired in parallel, allowing feasible volumetric imaging speed for human retina imaging. With the capability of imaging incoherent scattering contrast in 3D, the CoSLO revealed unique human retina features in 3D.

Three-dimensional motion-free imaging of retinal diseases is investigated. The Lissajous scan with the 3-mm scan width is repeated during the gradually shifting center of the scanning pattern. The lateral and axial motion amounts are estimated by using image correlation of OCT and OCTA en face images and the OCT intensity cross-sections, respectively. It results in an imaging area of 6.8 mm or more in diameter. Several eyes with retinal diseases have been scanned. The three-dimensional OCT images are well reconstructed without significant motion artifacts. Although some eyes exhibit severe eye motion, the three-dimensional structure of the retina is well reconstructed.

Neurodegenerative diseases, such as multiple sclerosis (MS), negatively impact the aging global population. MS damage various parts of the central nervous system, leading to various eye-movement abnormalities. We have built a retinal FreezEye Tracker (FET) to measure a wide dynamic range of eye movements of up to 10 degrees with an ultrahigh temporal and spatial resolution during visual tasks, including fixations, experiments with saccades, and smooth pursuit. To compare, we performed the same experiments with pupil tracker EyeLink 1000. The amplitudes of detected saccades are similar in both devices, but FET provides high-resolution details on eye trajectory during fixation periods.

Optical coherence tomography (OCT) allows for in vivo imaging of the individual layers of the retina, but conventional methods do not differentiate the outer nuclear layer (ONL) from Henle’s fiber layer (HFL). We have developed a robotically aligned OCT (RAOCT) system capable of automatically acquiring volumes at multiple pupil entry positions to reconstruct a retinal volume that distinguishes the HFL from the ONL. We utilized the RAOCT system to acquire such volumes from consented subjects, and used mosaicking to combine them into complete retinal volumes that show enhanced contrast in the HFL on all sides of the fovea.

Ophthalmic microsurgery involves the manipulation of thin, semi-transparent structures and has traditionally been performed using stereoscopic microscopes that provide an en face view of the surgical field. However, new therapeutic interventions such as subretinal injections require precise tool placement and dosing that are difficult to determine from the traditional microscope view. Optical coherence tomography (OCT) provides micron scale cross-sectional imaging and has become a gold standard in clinical ophthalmology settings, but its use in surgery has been more limited. The high-speed 400 kHz intraoperative system presented here provides valuable image guidance and quantitative metrics for a variety of human surgeries.

Unlike conventional ophthalmic OCT, robotically aligning OCT (RAOCT) removes the requirement for close patient/operator proximity and enables remote patient imaging by autonomously aligning itself to the patient while the operator is physically elsewhere. We report kilometer-scale distance between OCT operator and patient and the first robotically aligned OCT angiography images. We acquired remote volumetric RAOCT retinal images from healthy and diseased eyes at the Duke Eye Center on both room-to-room (10m between imager and subject) and between clinic sites(>10km between imager and subject). This can serve as a foundation for socially distanced or telehealth retinal OCT without physically present technicians.

The ability to perform multi-meridian, simultaneous OCT measurements of air-induced corneal deformation is expected to highly improve the accuracy of assessing corneal biomechanics. We propose a simplified method targeting 3-D deformation measurement that could be introduced to swept-source OCT systems. We utilize a spatial-depth-encoded multiplexing to provide a 9-spot measurement of the deformation. The method is promising for the assessment of corneal asymmetries and diagnosis of corneal pathologies such as keratoconus. We present in detail the system and key requirements to provide simultaneous 9-spot deformation measurement. Finally, results on porcine eyes ex vivo and human eye in vivo are presented.

We present the first clinical uses of time-domain FF-OCT in Anterior eye. Four patients with different eye pathologies (keratoconus, Fuch's endothelial dystrophy, age-related changes, post-PRK surgery) were imaged as part of the routine clinical examination (10 min per patient). FF-OCT resolved micron–size pathologies: striae (mechanical folds of corneal stroma), guttata (excrescences in Descemet's membrane), loss of endothelial cells and stromal cuts following the surgery. High resolution (1.7 µm) makes FF-OCT a promising tool for diagnosis of diseases at earlier stages than was possible before and their effective treatment with medication instead of surgery.

Mucin secretive conjunctival goblet cells (CGCs) play important roles for ocular surface homeostasis by forming the mucous layer of tear film, so CGC examination is important for diagnosis of various ocular surface diseases. Here we show that CGCs can be non-invasively imaged in real time and in high contrasts in animal models by moxifloxacin-based fluorescence microscopy (MBFM) using moxifloxacin antibiotic ophthalmic solution for specific CGC labeling. Newly developed MBFM was applied to both disease mouse and rabbit models and it detected CGC damage and recovery via longitudinal imaging. These results showed that MBFM has potentials for non-invasive CGC examination.

The focusing power of the human eye is provided by the cornea and the crystalline lens (CL). The latter provides the capacity to alter its shape and provide fine focus adjustment during the accommodation process. We extended standard structural OCT analysis to precisely follow dynamically subtle OCT signal phase changes within the human CL during accommodation. Imaging was performed using the SS-OCT system equipped with the Badal system to provide accommodation demand. The presented phase-sensitive analysis allows for distinct motion extraction such as lens wobbling, bulging, or any other deformation that can be significant for crystalline lens accommodation-associated biomechanics.

Mucin secretive conjunctival goblet cells (CGCs) in the eye are important for tear film stability and ocular surface health. Because CGC dysfunction is associated with various ocular surface diseases, non-invasive CGC examination will be of great help in the diagnosis and treatment. In this study, we developed a high-speed moxifloxacin-based extended depth-of-field microscopy for real-time CGC examination. The performance was demonstrated by high-speed CGC imaging of both mouse and rabbit models, in vivo. The imaging was insensitive to breathing motion, and the image resolution was sufficient to resolve individual CGCs in rabbit models.

The miniaturization of optical coherence tomography (OCT) systems could open up potential new markets, such as point-of-care application, home OCT to regularly monitor disease and treatment progress, and in low-resource settings. Photonic integrated circuits (PIC) are considered an attractive approach to miniaturize OCT. We present our recent achievements in in vivo retinal imaging with a PIC-based Mach-Zehnder interferometer integrated in a state-of-the-art ophthalmic OCT system. The system achieves 94 dB at 750 µW on the sample, running at 50 kHz. Preliminary results of a fully packaged 4-channel opto-electronic OCT engine further demonstrate the potential of PIC-based OCT.

We present a portable confocal scanning laser ophthalmoscope (cSLO) for high-resolution, widefield retinal imaging that uses a low-backreflection double-clad fiber for both illumination and collection to maintain robust alignment of confocal pinholes in a handheld form factor. The cSLO uses a new type of "hybrid spiral" scan pattern that is a hybridization of a constant angular velocity spiral with a constant linear velocity spiral. The hybrid spiral offers efficiency advantages over traditional raster scan patterns while serving as a natural fixation target. Feasibility of the system is demonstrated by imaging a human volunteer.

The study of choroidal blood flow is severely limited by the deficiencies of existing flow imaging methods. We introduce a new framework and acquisition protocol for optical coherence tomography (OCT) flowmetry in the choroid. Our approach quantifies choroidal flow by applying a robust mathematical analysis to signals that are dynamically forward scattered (DFS) by choroidal vessels and reflected from static scatterers in the sclera. This DFS approach provides robust and quantitative flow measurements that are immune to angle and gradient artifacts. We further demonstrate a visualization of flow mapping in a healthy human eye.

Glaucoma is one of the leading causes of irreversible blindness worldwide. The disease causes the loss of retinal nerve fibers (RNFs) and therefore visual field impairment. It is thought that in the early stages of glaucoma, before the loss of RNFs, the intracellular microtubules of the RNF axons disappear, altering the birefringent properties of the RNF-Layer (RNFL). In this study, we measured the birefringent properties of the RNFL using a polarization-sensitive OCT to compare glaucoma vs. healthy eyes. We observed a significantly reduced birefringence of the RNFL in glaucoma compared to healthy controls (0.090 ±0.009°/µm vs. 0.100 ±0.010°/µm, p<0.001).

The depolarization capabilities of the retinal pigment epithelium (RPE) are known to be dependent on the position at the macula and may be caused or influenced by the concentration of melanin and lipofuscin. In this work, the depolarization distribution of the RPE in a large group of healthy eyes is investigated using polarization-sensitive OCT by calculating the degree of polarization uniformity (DOPU) in a large field of view. The results are compared to diseased eyes (glaucoma, age-related macular degeneration), which might help to detect early pathologic changes of the RPE.

Glaucoma is a chronic optic neuropathy that severely damages the optic nerve head. Accurate information about the nerve fiber bundle (RNFB) trajectories in the retinal nerve fiber layer (RNFL) can improve earlier diagnosis and monitoring. Using polarization-sensitive (PS) OCT volume data we propose a fully automatic method of tracing RNFB trajectories. Preliminary analysis of automatically constructed RNFB traces shows a higher concordance to manually performed traces in comparison to a purely mathematical model. On repeated measurements of the same eye they also provide a higher reproducibility than results delivered by different graders.

This Conference Presentation, “Pascal Rol award presentation,” was recorded for the Photonics West 2022 On-Demand.

Delivery of pharmaceuticals to the eye posterior poses a major challenge in ophthalmology. A promising drug delivery platform is indocyanine green (ICG) liposomes, which absorb near-infrared light resulting in a release of pharmaceutical molecules. The Modulight ophthalmic laser platform has been designed for treatments targeting the eye posterior and can also excite the absorption band of the liposomes. The laser connects with Modulight Cloud, enabling artificial intelligence (AI) based treatment planning by correlating the treatment parameters and success, which could increase the efficacy of future treatments. The same algorithms could deduce which treatment parameters work with which liposomal delivery parameters.