Weakly supervised individual ganglion cell segmentation from adaptive optics OCT images for glaucomatous damage assessment

Scale-Adaptive viable tumor burden estimation via histopathological microscopy image segmentation

Cancer segmentation in whole-slide images is a fundamental step for estimating tumor burden, which is crucial for cancer assessment. However, challenges such as vague boundaries and small regions dissociated from viable tumor areas make it a complex task. Considering the usefulness of multi-scale features in various vision-related tasks, we present a structure-aware, scale-adaptive feature selection method for efficient and accurate cancer segmentation. Built on a segmentation network with a popular encoder-decoder architecture, a scale-adaptive module is proposed to select more robust features that better represent vague, non-rigid boundaries. Furthermore, a structural similarity metric is introduced to enhance tissue structure awareness and improve small region segmentation. Additionally, advanced designs, including several attention mechanisms and selective-kernel convolutions, are incorporated into the baseline network for comparative study purposes. Extensive experimental results demonstrate that the proposed structure-aware, scale-adaptive network achieves outstanding performance in liver cancer segmentation compared to the top submitted results in the PAIP 2019 challenge. Further evaluation of colorectal cancer segmentation shows that the scale-adaptive module either improves the baseline network or outperforms other advanced attention mechanism designs, particularly when considering the trade-off between efficiency and accuracy. The source code is publicly available at https://github.com/IMOP-lab/Scale-Adaptive-Net.

A scoping review of advancements in machine learning for glaucoma: current trends and future direction

Introduction

Machine learning technology has demonstrated significant potential in glaucoma research, particularly in early diagnosis, predicting disease progression, evaluating treatment responses, and developing personalized treatment strategies. The application of machine learning not only enhances the understanding of the pathological mechanism of glaucoma and optimizes the diagnostic process but also provides patients with accurate medical services.

Methods

This study aimed to describe the current state of research, highlight directions for further development, and identify potential trends for improvement. This review was conducted following the scoping review of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension to showcase advancements in the application of machine learning in glaucoma research and treatment.

Results

We employed a comprehensive search strategy to retrieve literature from the Web of Science Core Collection database, ultimately including 3,581 articles in the analysis. Through data analysis, we identified current research hotspots, noted differences in researchers' attitudes and opinions, and predicted potential future development trends.

Discussion

We divided the research topics into six categories, clearly identifying "eye diseases", "retinal fundus imaging" and "risk factors" as the key terms for the development of this field. These findings signify the promising prospects of machine learning, particularly when integrated with multimodal technologies and large language models, to enhance the diagnosis and treatment of glaucoma.

Gray Frequency-Based Methodology for Assessing Cell Damage

Cell biology techniques offer a solid foundation for evaluating and forecasting the danger of pollutants in the investigations of environmental toxicology. Studies on ecological toxicity, medication development, and illness diagnosis depend on evaluating cellular damage. The morphology of stimulated cells can alter the light scattering and reflection, and the brightness of microscopic images of the cells. This study demonstrated that stimulation-damaged and normal cells had distinct gray value distributions which led to the proposal of a novel theory to measure cellular damage by image brightness. Second, various cell types were used to confirm the method's applicability. Additionally, an evaluation technique based on gray frequency analysis can be created to determine the extent of cellular damage. This approach provides an effective and helpful tool for cellular damage visualization and quantitative evaluation in environmental toxicity assessment.

Identifying retinal pigment epithelium cells in adaptive optics-optical coherence tomography images with partial annotations and superhuman accuracy

Retinal pigment epithelium (RPE) cells are essential for normal retinal function. Morphological defects in these cells are associated with a number of retinal neurodegenerative diseases. Owing to the cellular resolution and depth-sectioning capabilities, individual RPE cells can be visualized in vivo with adaptive optics-optical coherence tomography (AO-OCT). Rapid, cost-efficient, and objective quantification of the RPE mosaic's structural properties necessitates the development of an automated cell segmentation algorithm. This paper presents a deep learning-based method with partial annotation training for detecting RPE cells in AO-OCT images with accuracy better than human performance. We have made the code, imaging datasets, and the manual expert labels available online.

The AI revolution in glaucoma: Bridging challenges with opportunities

Recent advancements in artificial intelligence (AI) herald transformative potentials for reshaping glaucoma clinical management, improving screening efficacy, sharpening diagnosis precision, and refining the detection of disease progression. However, incorporating AI into healthcare usages faces significant hurdles in terms of developing algorithms and putting them into practice. When creating algorithms, issues arise due to the intensive effort required to label data, inconsistent diagnostic standards, and a lack of thorough testing, which often limits the algorithms' widespread applicability. Additionally, the "black box" nature of AI algorithms may cause doctors to be wary or skeptical. When it comes to using these tools, challenges include dealing with lower-quality images in real situations and the systems' limited ability to work well with diverse ethnic groups and different diagnostic equipment. Looking ahead, new developments aim to protect data privacy through federated learning paradigms, improving algorithm generalizability by diversifying input data modalities, and augmenting datasets with synthetic imagery. The integration of smartphones appears promising for using AI algorithms in both clinical and non-clinical settings. Furthermore, bringing in large language models (LLMs) to act as interactive tool in medicine may signify a significant change in how healthcare will be delivered in the future. By navigating through these challenges and leveraging on these as opportunities, the field of glaucoma AI will not only have improved algorithmic accuracy and optimized data integration but also a paradigmatic shift towards enhanced clinical acceptance and a transformative improvement in glaucoma care.

VSG-GAN: A high-fidelity image synthesis method with semantic manipulation in retinal fundus image

In recent years, advancements in retinal image analysis, driven by machine learning and deep learning techniques, have enhanced disease detection and diagnosis through automated feature extraction. However, challenges persist, including limited data set diversity due to privacy concerns and imbalanced sample pairs, hindering effective model training. To address these issues, we introduce the vessel and style guided generative adversarial network (VSG-GAN), an innovative algorithm building upon the foundational concept of GAN. In VSG-GAN, a generator and discriminator engage in an adversarial process to produce realistic retinal images. Our approach decouples retinal image generation into distinct modules: the vascular skeleton and background style. Leveraging style transformation and GAN inversion, our proposed hierarchical variational autoencoder module generates retinal images with diverse morphological traits. In addition, the spatially adaptive denormalization module ensures consistency between input and generated images. We evaluate our model on MESSIDOR and RITE data sets using various metrics, including structural similarity index measure, inception score, Fréchet inception distance, and kernel inception distance. Our results demonstrate the superiority of VSG-GAN, outperforming existing methods across all evaluation assessments. This underscores its effectiveness in addressing data set limitations and imbalances. Our algorithm provides a novel solution to challenges in retinal image analysis by offering diverse and realistic retinal image generation. Implementing the VSG-GAN augmentation approach on downstream diabetic retinopathy classification tasks has shown enhanced disease diagnosis accuracy, further advancing the utility of machine learning in this domain.

Characterizing Presumed Displaced Retinal Ganglion Cells in the Living Human Retina of Healthy and Glaucomatous Eyes

Purpose

The purpose of this study was to investigate the large somas presumed to be displaced retinal ganglion cells (dRGCs) located in the inner nuclear layer (INL) of the living human retina. Whereas dRGCs have previously been studied in mammals and human donor tissue, they have never been investigated in the living human retina.

Methods

Five young, healthy subjects and three subjects with varying types of glaucoma were imaged at multiple locations in the macula using adaptive optics optical coherence tomography. In the acquired volumes, bright large somas at the INL border with the inner plexiform layer were identified, and the morphometric biomarkers of soma density, en face diameter, and spatial distribution were measured at up to 13 degrees retinal eccentricity. Susceptibility to glaucoma was assessed.

Results

In the young, healthy individuals, mean density of the bright, large somas was greatest foveally (550 and 543 cells/mm2 at 2 degrees temporal and nasal, respectively) and decreased with increasing retinal eccentricity (38 cells/mm2 at 13 degrees temporal, the farthest we measured). Soma size distribution showed the opposite trend with diameters and size variation increasing with retinal eccentricity, from 12.7 ± 1.8 µm at 2 degrees to 15.7 ± 3.5 µm at 13 degrees temporal, and showed evidence of a bimodal distribution in more peripheral locations. Within and adjacent to the arcuate defects of the subjects with glaucoma, density of the bright large somas was significantly lower than found in the young, healthy individuals.

Conclusions

Our results suggest that the bright, large somas at the INL border are likely comprised of dRGCs but amacrine cells may contribute too. These somas appear highly susceptible to glaucomatous damage.

Reducing manual labeling requirements and improved retinal ganglion cell identification in 3D AO-OCT volumes using semi-supervised learning

Adaptive optics-optical coherence tomography (AO-OCT) allows for the three-dimensional visualization of retinal ganglion cells (RGCs) in the living human eye. Quantitative analyses of RGCs have significant potential for improving the diagnosis and monitoring of diseases such as glaucoma. Recent advances in machine learning (ML) have made possible the automatic identification and analysis of RGCs within the complex three-dimensional retinal volumes obtained with such imaging. However, the current state-of-the-art ML approach relies on fully supervised training, which demands large amounts of training labels. Each volume requires many hours of expert manual annotation. Here, two semi-supervised training schemes are introduced, (i) cross-consistency training and (ii) cross pseudo supervision that utilize unlabeled AO-OCT volumes together with a minimal set of labels, vastly reducing the labeling demands. Moreover, these methods outperformed their fully supervised counterpart and achieved accuracy comparable to that of human experts.

In vivo imaging of human retinal ganglion cells using optical coherence tomography without adaptive optics

Retinal ganglion cells play an important role in human vision, and their degeneration results in glaucoma and other neurodegenerative diseases. Imaging these cells in the living human retina can greatly improve the diagnosis and treatment of glaucoma. However, owing to their translucent soma and tight packing arrangement within the ganglion cell layer (GCL), successful imaging has only been achieved with sophisticated research-grade adaptive optics (AO) systems. For the first time we demonstrate that GCL somas can be resolved and cell morphology can be quantified using non-AO optical coherence tomography (OCT) devices with optimal parameter configuration and post-processing.

Total variation regularized tensor ring decomposition for OCT image denoising and super-resolution

This paper suggests a novel hybrid tensor-ring (TR) decomposition and first-order tensor-based total variation (FOTTV) model, known as the TRFOTTV model, for super-resolution and noise suppression of optical coherence tomography (OCT) images. OCT imaging faces two fundamental problems undermining correct OCT-based diagnosis: significant noise levels and low sampling rates to speed up the capturing process. Inspired by the effectiveness of TR decomposition in analyzing complicated data structures, we suggest the TRFOTTV model for noise suppression and super-resolution of OCT images. Initially, we extract the nonlocal 3D patches from OCT data and group them to create a third-order low-rank tensor. Subsequently, using TR decomposition, we extract the correlations among all modes of the grouped OCT tensor. Finally, FOTTV is integrated into the TR model to enhance spatial smoothness in OCT images and conserve layer structures more effectively. The proximal alternating minimization and alternative direction method of multipliers are applied to solve the obtained optimization problem. The effectiveness of the suggested method is verified by four OCT datasets, demonstrating superior visual and numerical outcomes compared to state-of-the-art procedures.

Multifunctional adaptive optics optical coherence tomography allows cellular scale reflectometry, polarimetry, and angiography in the living human eye

Clinicians are unable to detect glaucoma until substantial loss or dysfunction of retinal ganglion cells occurs. To this end, novel measures are needed. We have developed an optical imaging solution based on adaptive optics optical coherence tomography (AO-OCT) to discern key clinical features of glaucoma and other neurodegenerative diseases at the cellular scale in the living eye. Here, we test the feasibility of measuring AO-OCT-based reflectance, retardance, optic axis orientation, and angiogram at specifically targeted locations in the living human retina and optic nerve head. Multifunctional imaging, combined with focus stacking and global image registration algorithms, allows us to visualize cellular details of retinal nerve fiber bundles, ganglion cell layer somas, glial septa, superior vascular complex capillaries, and connective tissues. These are key histologic features of neurodegenerative diseases, including glaucoma, that are now measurable in vivo with excellent repeatability and reproducibility. Incorporating this noninvasive cellular-scale imaging with objective measurements will significantly enhance existing clinical assessments, which is pivotal in facilitating the early detection of eye disease and understanding the mechanisms of neurodegeneration.

Imaging the Retinal Vascular Mural Cells In Vivo: Elucidating the Timeline of Their Loss in Diabetic Retinopathy

BACKGROUND

Vascular mural cells (VMCs) are integral components of the retinal vasculature with critical homeostatic functions such as maintaining the inner blood-retinal barrier and vascular tone, as well as supporting the endothelial cells. Histopathologic donor eye studies have shown widespread loss of pericytes and smooth muscle cells, the 2 main VMC types, suggesting these cells are critical to the pathogenesis of diabetic retinopathy (DR). There remain, however, critical gaps in our knowledge regarding the timeline of VMC demise in human DR.

METHODS

In this study, we address this gap using adaptive optics scanning laser ophthalmoscopy to quantify retinal VMC density in eyes with no retinal disease (healthy), subjects with diabetes without diabetic retinopathy, and those with clinical DR and diabetic macular edema. We also used optical coherence tomography angiography to quantify capillary density of the superficial and deep capillary plexuses in these eyes.

RESULTS

Our results indicate significant VMC loss in retinal arterioles before the appearance of classic clinical signs of DR (diabetes without diabetic retinopathy versus healthy, 5.0±2.0 versus 6.5±2.0 smooth muscle cells per 100 µm; P<0.05), while a significant reduction in capillary VMC density (5.1±2.3 in diabetic macular edema versus 14.9±6.0 pericytes per 100 µm in diabetes without diabetic retinopathy; P=0.01) and capillary density (superficial capillary plexus vessel density, 37.6±3.8 in diabetic macular edema versus 45.5±2.4 in diabetes without diabetic retinopathy; P<0.0001) is associated with more advanced stages of clinical DR, particularly diabetic macular edema.

CONCLUSIONS

Our results offer a new framework for understanding the pathophysiologic course of VMC compromise in DR, which may facilitate the development and monitoring of therapeutic strategies aimed at VMC preservation and potentially the prevention of clinical DR and its associated morbidity. Imaging retinal VMCs provides an unparalleled opportunity to visualize these cells in vivo and may have wider implications in a range of diseases where these cells are disrupted.

Is Primary Open-Angle Glaucoma a Vascular Disease? Assessment of the Relationship between Retinal Arteriolar Morphology and Glaucoma Severity Using Adaptive Optics

BACKGROUND

Retinal vascular abnormalities may be associated with glaucomatous damage. Adaptive optics (AO) is a new technology that enables the analysis of retinal vasculature at the cellular level in vivo. The purpose of this study was to evaluate retinal arteriolar parameters using the rtx1 adaptive optics fundus camera (AO-FC) in patients with primary open-angle glaucoma (POAG) at different stages and to investigate the relationship between these parameters and changes in spectral-domain optical coherence tomography (SD-OCT) and perimetry.

METHODS

Parameters of the retinal supratemporal and infratemporal arterioles (wall thickness (WT), lumen diameter (LD), total diameter (TD), wall-to-lumen ratio (WLR), and cross-sectional area of the vascular wall (WCSA)) were analysed with the rtx1 in 111 POAG eyes, which were divided into three groups according to the severity of the disease, and 70 healthy eyes. The associations between RTX1 values and the cup-to-disk ratio, SD-OCT parameters, and visual field parameters were assessed.

RESULTS

Compared with the control group, the POAG groups showed significantly smaller TD and LD values (p < 0.05) and significantly higher WLR and WT values (p < 0.05) for the supratemporal and infratemporal arterioles. TD was significantly positively correlated with the retinal nerve fibre layer (RNFL) and ganglion cell complex (GCC) (p < 0.05). LD was significantly positively correlated with the RNFL, GCC, and rim area (p < 0.05). The WLR was significantly negatively correlated with the RNFL, GCC, rim area, and MD (p < 0.05), while it was significantly positively correlated with the cup-to-disc ratio and PSD (p < 0.05).

CONCLUSIONS

The results suggest that vascular dysfunction is present in POAG, even at a very early stage of glaucoma, and increases with the severity of the disease.

Cellular-Level Visualization of Retinal Pathology in Multiple Sclerosis With Adaptive Optics

Purpose

To apply adaptive optics-optical coherence tomography (AO-OCT) to quantify multiple sclerosis (MS)-induced changes in axonal bundles in the macular nerve fiber layer, ganglion cell somas, and macrophage-like cells at the vitreomacular interface.

Methods

We used AO-OCT imaging in a pilot study of MS participants (n = 10), including those without and with a history of optic neuritis (ON, n = 4), and healthy volunteers (HV, n = 9) to reveal pathologic changes to inner retinal cells and structures affected by MS.

Results

We found that nerve fiber layer axonal bundles had 38% lower volume in MS participants (1.5 × 10-3 mm3) compared to HVs (2.4 × 10-3 mm3; P < 0.001). Retinal ganglion cell (RGC) density was 51% lower in MS participants (12.3 cells/mm2 × 1000) compared to HVs (25.0 cells/mm2 × 1000; P < 0.001). Spatial differences across the macula were observed in RGC density. RGC diameter was 15% higher in MS participants (11.7 µm) compared to HVs (10.1 µm; P < 0.001). A nonsignificant trend of higher density of macrophage-like cells in MS eyes was also observed. For all AO-OCT measures, outcomes were worse for MS participants with a history of ON compared to MS participants without a history of ON. AO-OCT measures were associated with key visual and physical disabilities in the MS cohort.

Conclusions

Our findings demonstrate the utility of AO-OCT for highly sensitive and specific detection of neurodegenerative changes in MS. Moreover, the results shed light on the mechanisms that underpin specific neuronal pathology that occurs when MS attacks the retina. The new findings support the further development of AO-based biomarkers for MS.

Unraveling the complexity of Optical Coherence Tomography image segmentation using machine and deep learning techniques: A review

Optical Coherence Tomography (OCT) is an emerging technology that provides three-dimensional images of the microanatomy of biological tissue in-vivo and at micrometer-scale resolution. OCT imaging has been widely used to diagnose and manage various medical diseases, such as macular degeneration, glaucoma, and coronary artery disease. Despite its wide range of applications, the segmentation of OCT images remains difficult due to the complexity of tissue structures and the presence of artifacts. In recent years, different approaches have been used for OCT image segmentation, such as intensity-based, region-based, and deep learning-based methods. This paper reviews the major advances in state-of-the-art OCT image segmentation techniques. It provides an overview of the advantages and limitations of each method and presents the most relevant research works related to OCT image segmentation. It also provides an overview of existing datasets and discusses potential clinical applications. Additionally, this review gives an in-depth analysis of machine learning and deep learning approaches for OCT image segmentation. It outlines challenges and opportunities for further research in this field.

Derivation of human retinal cell densities using high-density, spatially localized optical coherence tomography data from the human retina

This study sought to identify demographic variations in retinal thickness measurements from optical coherence tomography (OCT), to enable the calculation of cell density parameters across the neural layers of the healthy human macula. From macular OCTs (n = 247), ganglion cell (GCL), inner nuclear (INL), and inner segment-outer segment (ISOS) layer measurements were extracted using a customized high-density grid. Variations with age, sex, ethnicity, and refractive error were assessed with multiple linear regression analyses, with age-related distributions further assessed using hierarchical cluster analysis and regression models. Models were tested on a naïve healthy cohort (n = 40) with Mann-Whitney tests to determine generalizability. Quantitative cell density data were calculated from histological data from previous human studies. Eccentricity-dependent variations in OCT retinal thickness closely resemble topographic cell density maps from human histological studies. Age was consistently identified as significantly impacting retinal thickness (p = .0006, .0007, and .003 for GCL, INL and ISOS), with gender affecting ISOS only (p < .0001). Regression models demonstrated that age-related changes in the GCL and INL begin in the 30th decade and were linear for the ISOS. Model testing revealed significant differences in INL and ISOS thickness (p = .0008 and .0001; however, differences fell within the OCT's axial resolution. Qualitative comparisons show close alignment between OCT and histological cell densities when using unique, high-resolution OCT data, and correction for demographics-related variability. Overall, this study describes a process to calculate in vivo cell density from OCT for all neural layers of the human retina, providing a framework for basic science and clinical investigations.

The application of artificial intelligence in glaucoma diagnosis and prediction

Artificial intelligence is a multidisciplinary and collaborative science, the ability of deep learning for image feature extraction and processing gives it a unique advantage in dealing with problems in ophthalmology. The deep learning system can assist ophthalmologists in diagnosing characteristic fundus lesions in glaucoma, such as retinal nerve fiber layer defects, optic nerve head damage, optic disc hemorrhage, etc. Early detection of these lesions can help delay structural damage, protect visual function, and reduce visual field damage. The development of deep learning led to the emergence of deep convolutional neural networks, which are pushing the integration of artificial intelligence with testing devices such as visual field meters, fundus imaging and optical coherence tomography to drive more rapid advances in clinical glaucoma diagnosis and prediction techniques. This article details advances in artificial intelligence combined with visual field, fundus photography, and optical coherence tomography in the field of glaucoma diagnosis and prediction, some of which are familiar and some not widely known. Then it further explores the challenges at this stage and the prospects for future clinical applications. In the future, the deep cooperation between artificial intelligence and medical technology will make the datasets and clinical application rules more standardized, and glaucoma diagnosis and prediction tools will be simplified in a single direction, which will benefit multiple ethnic groups.

Deep learning-enabled volumetric cone photoreceptor segmentation in adaptive optics optical coherence tomography images of normal and diseased eyes

Objective quantification of photoreceptor cell morphology, such as cell diameter and outer segment length, is crucial for early, accurate, and sensitive diagnosis and prognosis of retinal neurodegenerative diseases. Adaptive optics optical coherence tomography (AO-OCT) provides three-dimensional (3-D) visualization of photoreceptor cells in the living human eye. The current gold standard for extracting cell morphology from AO-OCT images involves the tedious process of 2-D manual marking. To automate this process and extend to 3-D analysis of the volumetric data, we propose a comprehensive deep learning framework to segment individual cone cells in AO-OCT scans. Our automated method achieved human-level performance in assessing cone photoreceptors of healthy and diseased participants captured with three different AO-OCT systems representing two different types of point scanning OCT: spectral domain and swept source.

In vivo chromatic and spatial tuning of foveolar retinal ganglion cells in Macaca fascicularis

The primate fovea is specialized for high acuity chromatic vision, with the highest density of cone photoreceptors and a disproportionately large representation in visual cortex. The unique visual properties conferred by the fovea are conveyed to the brain by retinal ganglion cells, the somas of which lie at the margin of the foveal pit. Microelectrode recordings of these centermost retinal ganglion cells have been challenging due to the fragility of the fovea in the excised retina. Here we overcome this challenge by combining high resolution fluorescence adaptive optics ophthalmoscopy with calcium imaging to optically record functional responses of foveal retinal ganglion cells in the living eye. We use this approach to study the chromatic responses and spatial transfer functions of retinal ganglion cells using spatially uniform fields modulated in different directions in color space and monochromatic drifting gratings. We recorded from over 350 cells across three Macaca fascicularis primates over a time period of weeks to months. We find that the majority of the L vs. M cone opponent cells serving the most central foveolar cones have spatial transfer functions that peak at high spatial frequencies (20-40 c/deg), reflecting strong surround inhibition that sacrifices sensitivity at low spatial frequencies but preserves the transmission of fine detail in the retinal image. In addition, we fit to the drifting grating data a detailed model of how ganglion cell responses draw on the cone mosaic to derive receptive field properties of L vs. M cone opponent cells at the very center of the foveola. The fits are consistent with the hypothesis that foveal midget ganglion cells are specialized to preserve information at the resolution of the cone mosaic. By characterizing the functional properties of retinal ganglion cells in vivo through adaptive optics, we characterize the response characteristics of these cells in situ.

Deep learning and optical coherence tomography in glaucoma: Bridging the diagnostic gap on structural imaging

Glaucoma is a leading cause of progressive blindness and visual impairment worldwide. Microstructural evidence of glaucomatous damage to the optic nerve head and associated tissues can be visualized using optical coherence tomography (OCT). In recent years, development of novel deep learning (DL) algorithms has led to innovative advances and improvements in automated detection of glaucomatous damage and progression on OCT imaging. DL algorithms have also been trained utilizing OCT data to improve detection of glaucomatous damage on fundus photography, thus improving the potential utility of color photos which can be more easily collected in a wider range of clinical and screening settings. This review highlights ten years of contributions to glaucoma detection through advances in deep learning models trained utilizing OCT structural data and posits future directions for translation of these discoveries into the field of aging and the basic sciences.

Automated segmentation of the ciliary muscle in OCT images using fully convolutional networks

Quantifying shape changes in the ciliary muscle during accommodation is essential in understanding the potential role of the ciliary muscle in presbyopia. The ciliary muscle can be imaged in-vivo using OCT but quantifying the ciliary muscle shape from these images has been challenging both due to the low contrast of the images at the apex of the ciliary muscle and the tedious work of segmenting the ciliary muscle shape. We present an automatic-segmentation tool for OCT images of the ciliary muscle using fully convolutional networks. A study using a dataset of 1,039 images shows that the trained fully convolutional network can successfully segment ciliary muscle images and quantify ciliary muscle thickness changes during accommodation. The study also shows that EfficientNet outperforms other current backbones of the literature.

Spatiotemporal absorption fluctuation imaging based on U-Net

SIGNIFICANCE

Full-field optical angiography is critical for vascular disease research and clinical diagnosis. Existing methods struggle to improve the temporal and spatial resolutions simultaneously.

AIM

Spatiotemporal absorption fluctuation imaging (ST-AFI) is proposed to achieve dynamic blood flow imaging with high spatial and temporal resolutions.

APPROACH

ST-AFI is a dynamic optical angiography based on a low-coherence imaging system and U-Net. The system was used to acquire a series of dynamic red blood cell (RBC) signals and static background tissue signals, and U-Net is used to predict optical absorption properties and spatiotemporal fluctuation information. U-Net was generally used in two-dimensional blood flow segmentation as an image processing algorithm for biomedical imaging. In the proposed approach, the network simultaneously analyzes the spatial absorption coefficient differences and the temporal dynamic absorption fluctuation.

RESULTS

The spatial resolution of ST-AFI is up to 4.33  μm, and the temporal resolution is up to 0.032 s. In vivo experiments on 2.5-day-old chicken embryos were conducted. The results demonstrate that intermittent RBCs flow in capillaries can be resolved, and the blood vessels without blood flow can be suppressed.

CONCLUSIONS

Using ST-AFI to achieve convolutional neural network (CNN)-based dynamic angiography is a novel approach that may be useful for several clinical applications. Owing to their strong feature extraction ability, CNNs exhibit the potential to be expanded to other blood flow imaging methods for the prediction of the spatiotemporal optical properties with improved temporal and spatial resolutions.

Cell - Vessel Mismatch in Glaucoma: Correlation of Ganglion Cell Layer Soma and Capillary Densities

Purpose

The purpose of this study was to characterize the relationship between retinal ganglion cell layer (GCL) soma density and capillary density in glaucomatous eyes.

Methods

Six glaucoma subjects with known hemifield defects and 6 age-matched controls were imaged with adaptive optics - optical coherence tomography (AO-OCT) at 6 locations: 3 degrees, 6 degrees, and 12 degrees temporal to the fovea above and below the midline. GCL soma density and capillary density were measured at each location. Coefficients of determination (pseudo R2) and slopes between GCL soma and capillary density were determined from mixed-effects regressions and were compared between glaucoma and control subjects, between more and less affected hemifield in subjects with glaucoma, and between subjects with early and moderate glaucoma, both in a local, bivariate model and then a global, multivariable model controlling for eccentricity and soma size.

Results

The global correlation between GCL soma and capillary density was stronger in control versus subjects with glaucoma (R2 = 0.59 vs. 0.22), less versus more affected hemifields (R2 = 0.55 vs. 0.01), and subjects with early versus moderate glaucoma subjects (R2 = 0.44 vs. 0.18). When controlling for eccentricity and soma size, we noted an inverse soma-capillary density local relationship in subjects with glaucoma (-388 ± 190 cells/mm2 per 1% change in capillary density, P = 0.046) and more affected hemifields (-602 ± 257 cells/mm2 per 1% change in capillary density, P = 0.03).

Conclusions

An inverted soma-capillary density local relationship in areas affected by glaucoma potentially explains weaker global correlations observed between GCL soma and capillary density, suggesting cell-vessel mismatch is associated with the disease.

Weakly supervised individual ganglion cell segmentation from adaptive optics OCT images for glaucomatous damage assessment

Cell-level quantitative features of retinal ganglion cells (GCs) are potentially important biomarkers for improved diagnosis and treatment monitoring of neurodegenerative diseases such as glaucoma, Parkinson's disease, and Alzheimer's disease. Yet, due to limited resolution, individual GCs cannot be visualized by commonly used ophthalmic imaging systems, including optical coherence tomography (OCT), and assessment is limited to gross layer thickness analysis. Adaptive optics OCT (AO-OCT) enables in vivo imaging of individual retinal GCs. We present an automated segmentation of GC layer (GCL) somas from AO-OCT volumes based on weakly supervised deep learning (named WeakGCSeg), which effectively utilizes weak annotations in the training process. Experimental results show that WeakGCSeg is on par with or superior to human experts and is superior to other state-of-the-art networks. The automated quantitative features of individual GCLs show an increase in structure-function correlation in glaucoma subjects compared to using thickness measures from OCT images. Our results suggest that by automatic quantification of GC morphology, WeakGCSeg can potentially alleviate a major bottleneck in using AO-OCT for vision research.