Search

Visible light optical coherence tomography (vis-OCT) is gaining traction for retinal imaging due to its high resolution and functional capabilities. However, the significant absorption of hemoglobin in the visible light range leads to pronounced shadow artifacts from retinal blood vessels, posing challenges for accurate layer segmentation. In this study, we present BreakNet, a multi-scale Transformer-based segmentation model designed to address boundary discontinuities caused by these shadow artifacts. BreakNet utilizes hierarchical Transformer and convolutional blocks to extract multi-scale global and local feature maps, capturing essential contextual, textural, and edge characteristics. The model incorporates decoder blocks that expand pathwaproys to enhance the extraction of fine details and semantic information, ensuring precise segmentation. Evaluated on rodent retinal images acquired with prototype vis-OCT, BreakNet demonstrated superior performance over state-of-the-art segmentation models, such as TCCT-BP and U-Net, even when faced with limited-quality ground truth data. Our findings indicate that BreakNet has the potential to significantly improve retinal quantification and analysis.

Fundus photography (FP) remains the primary imaging modality in screening various retinal diseases including age-related macular degeneration, diabetic retinopathy and glaucoma. FP allows the clinician to examine the ocular fundus structures such as the macula, the optic disc (OD) and retinal vessels, whose visibility and clarity in an FP image remain central to ensuring diagnostic accuracy, and hence determine the diagnostic quality (DQ). Images with low DQ, resulting from eye movement, improper illumination and other possible causes, should obviously be recaptured. However, the technician, often unfamiliar with DQ criteria, initiates recapture only based on expert feedback. The process potentially engages the imaging device multiple times for single subject, and wastes the time and effort of the ophthalmologist, the technician and the subject. The burden could be prohibitive in case of teleophthalmology, where obtaining feedback from the remote expert entails additional communication cost and delay. Accordingly, a strong need for automated diagnostic quality assessment (DQA) has been felt, where an image is immediately assigned a DQ category. In response, motivated by the notional continuum of DQ, we propose a hierarchical deep learning (DL) architecture to distinguish between good, usable and unusable categories. On the public EyeQ dataset, we achieve an accuracy of 89.44%, improving upon existing methods. In addition, using gradient based class activation map (Grad-CAM), we generate a visual explanation which agrees with the expert intuition. Future FP cameras equipped with the proposed DQA algorithm will potentially improve the efficacy of the teleophthalmology as well as the traditional system.

We describe the advantages of diffuse illumination in laser holography for ophthalmology. The presence of a diffusing element introduces an angular diversity of the optical radiation and reduces its spatial coherence, which spreads out the energy distribution of the illumination beam in the focal plane of the eyepiece. The field of view of digitally computed retinal images can easily be increased as the eyepiece can be moved closer to the cornea to obtain a Maxwellian view of the retina without compromising ocular safety. Compliance with American and European safety standards for ophthalmic devices is more easily obtained by preventing the presence of a laser hot spot observed in front of the cornea in the absence of a scattering element. Diffuse laser illumination does not introduce any adverse effects on digitally computed laser Doppler images., Comment: 9 pages

Laser Doppler holography (LDH) is a full-field blood flow imaging technique able to reveal human retinal and choroidal blood flow with high temporal resolution. We here report on using LDH in the anterior segment of the eye without making changes to the instrument. Blood flow in the bulbar conjunctiva and episclera as well as in corneal neovascularization can be effectively imaged. We additionally demonstrate simultaneous holographic imaging of the anterior and posterior segments by simply adapting the numerical propagation distance to the plane of interest. We used this feature to track the movements of the retina and pupil with high temporal resolution. Finally, we show that the light backscattered by the retina can be used for retro-illumination of the anterior segment. Hence digital holography can reveal opacities caused by absorption or diffusion in the cornea and eye lens.

Retinal degenerative diseases lead to the blindness of millions of people around the world. In case of age-related macular degeneration (AMD), the atrophy of retinal pigment epithelium (RPE) precedes neural dystrophy. But as crucial as understanding both healthy and pathological RPE cell physiology is for those diseases, no current technique allows subcellular in vivo or in vitro live observation of this critical cell layer. To fill this gap, we propose dynamic full-field OCT (D-FFOCT) as a candidate for live observation of in vitro RPE phenotype. In this way, we monitored primary porcine and human stem cell-derived RPE cells in stress model conditions by performing scratch assays. In this study, we quantified wound healing parameters on the stressed RPE, and observed different cell phenotypes, displayed by the D-FFOCT signal. In order to decipher the subcellular contributions to these dynamic profiles, we performed immunohistochemistry to identify which organelles generate the signal and found mitochondria to be the main contributor to D-FFOCT contrast. Altogether, D-FFOCT appears to be an innovative method to follow degenerative disease evolution and could be an appreciated method in the future for live patient diagnostics and to direct treatment choice.

Laser Doppler holography was introduced as a full-field imaging technique to measure blood flow in the retina and choroid with an as yet unrivaled temporal resolution. We here investigate separating the different contributions to the power Doppler signal in order to isolate the flow waveforms of vessels in the posterior pole of the human eye. Distinct flow behaviors are found in retinal arteries and veins with seemingly interrelated waveforms. We demonstrate a full field mapping of the local resistivity index, and the possibility to perform unambiguous identification of retinal arteries and veins on the basis of their systolodiastolic variations. Finally we investigate the arterial flow waveforms in the retina and choroid and find synchronous and similar waveforms, although with a lower pulsatility in choroidal vessels. This work demonstrates the potential held by laser Doppler holography to study ocular hemodynamics in healthy and diseased eyes.

The choroid is a highly vascularized tissues supplying the retinal pigment epithelium and photoreceptors. Its implication in retinal diseases is gaining increasing interest. However, investigating the anatomy and flow of the choroid remains challenging. Here we show that laser Doppler holography provides high contrast imaging of choroidal vessels in humans, with a spatial resolution comparable to state of the art indocyanine green angiography and optical coherence tomography. Additionally, laser Doppler holography contributes to sort out choroidal arteries and veins by using a power Doppler spectral analysis. We thus demonstrate the potential of laser Doppler holography to improve our understanding of the anatomy and flow of the choroidal vascular network.

Optical coherence tomography offers astounding opportunities to image the complex structure of living tissue, but lacks functional information. We present dynamic full-field optical coherence tomography to image living human induced pluripotent stem cell-derived retinal organoids non-invasively. Colored images with an endogenous contrast linked to organelle motility are generated, with sub-micrometer spatial resolution and millisecond temporal resolution, opening an avenue to identify specific cell types in living tissue via their function., Comment: 14 pages, 5 figures, 1 table, 6 videos

The majority of inherited retinal degenerations are due to photoreceptor cell death. In many cases ganglion cells are spared making it possible to stimulate them to restore visual function. Several studies (Bi et al., 2006; Lin et al., 2008; Sengupta et al., 2016; Caporale et al., 2011; Berry et al., 2017) have shown that it is possible to express an optogenetic protein in ganglion cells and make them light sensitive. This is a promising strategy to restore vision since optical targeting may be more precise than electrical stimulation with a retinal prothesis. However the spatial resolution of optogenetically-reactivated retinas has not been measured with fine-grained stimulation patterns. Since the optogenetic protein is also expressed in axons, it is unclear if these neurons will only be sensitive to the stimulation of a small region covering their somas and dendrites, or if they will also respond to any stimulation overlapping with their axon, dramatically impairing spatial resolution. Here we recorded responses of mouse and macaque retinas to random checkerboard patterns following an in vivo optogenetic therapy. We show that optogenetically activated ganglion cells are each sensitive to a small region of visual space. A simple model based on this small receptive field predicted accurately their responses to complex stimuli. From this model, we simulated how the entire population of light sensitive ganglion cells would respond to letters of different sizes. We then estimated the maximal acuity expected by a patient, assuming it could make an optimal use of the information delivered by this reactivated retina. The obtained acuity is above the limit of legal blindness. This high spatial resolution is a promising result for future clinical studies.

As the lateral resolution of FFOCT with spatially incoherent illumination has been shown to be insensitive to aberrations, we demonstrate high resolution en face full-field OCT (FFOCT) retinal imaging without wavefront correction in the human eye in vivo for the first time. A combination of FFOCT with spectral-domain OCT (SDOCT) is applied for real-time matching of the optical path lengths (OPL) of FFOCT. Through the real-time cross-sectional SDOCT images, the OPL of the FFOCT reference arm is matched with different retinal layers in the FFOCT sample arm. Thus, diffraction limited FFOCT images of multiple retinal layers are acquired at both the near periphery and the fovea. The en face FFOCT retinal images reveal information about various structures such as the nerve fiber orientation, the blood vessel distribution, and the photoreceptor mosaic.

We performed non-invasive video imaging of retinal blood flow in a pigmented rat by holographic interferometry of near-infrared laser light backscattered by retinal tissue, beating against an off-axis reference beam sampled at a frame rate of 39 kHz with a high throughput camera. Local Doppler contrasts emerged from the envelopes of short-time Fourier transforms and the phase of autocorrelation functions of holograms rendered by Fresnel transformation. This approach permitted imaging of blood flow in large retinal vessels (30 microns diameter) over 400 by 400 pixels with a spatial resolution of 8 microns and a temporal resolution of 6.5 ms., Comment: 4 pages, 5 figures, 4 links