DIRECTIONAL OPTICAL COHERENCE TOMOGRAPHY IMAGING OF MACULAR PATHOLOGY.

PURPOSE: To survey the impact of directional reflectivity on structures within optical coherence tomography images in retinal pathology. METHODS: Sets of commercial optical coherence tomography images taken from multiple pupil positions were analyzed. These directional optical coherence tomography sets revealed directionally reflective structures within the retina. After ensuring sufficient image quality, resulting hybrid and composite images were characterized by assessing the Henle fiber layer, outer nuclear layer, ellipsoid zone, and interdigitation zone. Additionally, hybrid images were reviewed for novel directionally reflective pathological features. RESULTS: Cross-sectional directional optical coherence tomography image sets were obtained in 75 eyes of 58 patients having a broad range of retinal pathologies. All cases showed improved visualization of the outer nuclear layer/Henle fiber layer interface, and outer nuclear layer thinning was, therefore, more apparent in several cases. The ellipsoid zone and interdigitation zone also demonstrated attenuation where a geometric impact of underlying pathology affected their orientation. Misdirected photoreceptors were also noted as a consistent direction-dependent change in ellipsoid zone reflectivity between regions of normal and absent ellipsoid zone. CONCLUSION: Directional optical coherence tomography enhances the understanding of retinal anatomy and pathology. This optical contrast yields more accurate identification of retinal structures and possible imaging biomarkers for photoreceptor-related pathology.

DIRECTIONAL OPTICAL COHERENCE TOMOGRAPHY PROVIDES ACCURATE OUTER NUCLEAR LAYER AND HENLE FIBER LAYER MEASUREMENTS.

PURPOSE: The outer nuclear layer (ONL) contains photoreceptor nuclei, and its thickness is an important biomarker for retinal degenerations. Accurate ONL thickness measurements are obscured in standard optical coherence tomography (OCT) images because of Henle fiber layer (HFL). Improved differentiation of the ONL and HFL boundary is made possible by using directional OCT, a method that purposefully varies the pupil entrance position of the OCT beam. METHODS: Fifty-seven normal eyes were imaged using multiple pupil entry positions with a commercial spectral domain OCT system. Cross-sectional image sets were registered to each other and segmented at the top of HFL, the border of HFL and the ONL and at the external limiting membrane. Thicknesses of the ONL and HFL were measured and analyzed. RESULTS: The true ONL and HFL thicknesses varied substantially by eccentricity and between individuals. The true macular ONL thickness comprised an average of 54.6% of measurements that also included HFL. The ONL and HFL thicknesses at specific retinal eccentricities were poorly correlated. CONCLUSION: Accurate ONL and HFL thickness measurements are made possible by the optical contrast of directional OCT. Distinguishing these individual layers can improve clinical trial endpoints and assessment of disease progression.

Estimating Global Visual Field Indices in Glaucoma by Combining Macula and Optic Disc OCT Scans Using 3-Dimensional Convolutional Neural Networks.

PURPOSE: To evaluate the accuracy at which visual field global indices could be estimated from OCT scans of the retina using deep neural networks and to quantify the contributions to the estimates by the macula (MAC) and the optic nerve head (ONH). DESIGN: Observational cohort study. PARTICIPANTS: A total of 10 370 eyes from 109 healthy patients, 697 glaucoma suspects, and 872 patients with glaucoma over multiple visits (median = 3). METHODS: Three-dimensional convolutional neural networks were trained to estimate global visual field indices derived from automated Humphrey perimetry (SITA 24-2) tests (Zeiss, Dublin, CA), using OCT scans centered on MAC, ONH, or both (MAC + ONH) as inputs. MAIN OUTCOME MEASURES: Spearman's rank correlation coefficients, Pearson's correlation coefficient, and absolute errors calculated for 2 indices: visual field index (VFI) and mean deviation (MD). RESULTS: The MAC + ONH achieved 0.76 Spearman's correlation coefficient and 0.87 Pearson's correlation for VFI and MD. Median absolute error was 2.7 for VFI and 1.57 decibels (dB) for MD. Separate MAC or ONH estimates were significantly less correlated and less accurate. Accuracy was dependent on the OCT signal strength and the stage of glaucoma severity. CONCLUSIONS: The accuracy of global visual field indices estimate is improved by integrating information from MAC and ONH in advanced glaucoma, suggesting that structural changes of the 2 regions have different time courses in the disease severity spectrum.

Directional Reflectivity of the Ellipsoid Zone in Dry Age-Related Macular Degeneration.

BACKGROUND AND OBJECTIVE: Ellipsoid zone (EZ) reflectivity on optical coherence tomography (OCT) is affected by the orientation of the scanning beam. The authors sought to determine how directional reflectivity changes in dry age-related macular degeneration (AMD). PATIENTS AND METHODS: Retrospective image analysis included 17 control and 20 dry AMD subjects. Directional OCT (D-OCT) was performed using multiple displaced pupil entrance positions. EZ pixel values and apparent incidence angles were measured. RESULTS: EZ reflectivity decreased in off-axis scans in controls (P < .001), AMD areas between drusen (P < .001), and AMD areas overlying drusen (P < .001). The magnitude of decrement in EZ reflectivity was significantly higher when incidence angles exceeded 10° in controls than in AMD areas between drusen (P = .024). CONCLUSION: EZ reflectivity in dry AMD may vary by incident angle of light less than in controls, possibly indicating alteration of photoreceptor orientation or integrity. [Ophthalmic Surg Lasers Imaging Retina. 2021;52:145-152.].

Attention-Guided 3D-CNN Framework for Glaucoma Detection and Structural-Functional Association Using Volumetric Images.

The direct analysis of 3D Optical Coherence Tomography (OCT) volumes enables deep learning models (DL) to learn spatial structural information and discover new bio-markers that are relevant to glaucoma. Downsampling 3D input volumes is the state-of-art solution to accommodate for the limited number of training volumes as well as the available computing resources. However, this limits the network's ability to learn from small retinal structures in OCT volumes. In this paper, our goal is to improve the performance by providing guidance to DL model during training in order to learn from finer ocular structures in 3D OCT volumes. Therefore, we propose an end-to-end attention guided 3D DL model for glaucoma detection and estimating visual function from retinal structures. The model consists of three pathways with the same network architecture but different inputs. One input is the original 3D-OCT cube and the other two are computed during training guided by the 3D gradient class activation heatmaps. Each pathway outputs the class-label and the whole model is trained concurrently to minimize the sum of losses from three pathways. The final output is obtained by fusing the predictions of the three pathways. Also, to explore the robustness and generalizability of the proposed model, we apply the model on a classification task for glaucoma detection as well as a regression task to estimate visual field index (VFI) (a value between 0 and 100). A 5-fold cross-validation with a total of 3782 and 10,370 OCT scans is used to train and evaluate the classification and regression models, respectively. The glaucoma detection model achieved an area under the curve (AUC) of 93.8% compared with 86.8% for a baseline model without the attention-guided component. The model also outperformed six different feature based machine learning approaches that use scanner computed measurements for training. Further, we also assessed the contribution of different retinal layers that are relevant to glaucoma. The VFI estimation model achieved a Pearson correlation and median absolute error of 0.75 and 3.6%, respectively, for a test set of size 3100 cubes.

Layer boundary evolution method for macular OCT layer segmentation.

Optical coherence tomography (OCT) is used to produce high resolution depth images of the retina and is now the standard of care for in-vivo ophthalmological assessment. It is also increasingly being used for evaluation of neurological disorders such as multiple sclerosis (MS). Automatic segmentation methods identify the retinal layers of the macular cube providing consistent results without intra- and inter-rater variation and is faster than manual segmentation. In this paper, we propose a fast multi-layer macular OCT segmentation method based on a fast level set method. Our framework uses contours in an optimized approach specifically for OCT layer segmentation over the whole macular cube. Our algorithm takes boundary probability maps from a trained random forest and iteratively refines the prediction to subvoxel precision. Evaluation on both healthy and multiple sclerosis subjects shows that our method is statistically better than a state-of-the-art graph-based method.

Retinal optical coherence tomography image enhancement via deep learning.

Optical coherence tomography (OCT) images of the retina are a powerful tool for diagnosing and monitoring eye disease. However, they are plagued by speckle noise, which reduces image quality and reliability of assessment. This paper introduces a novel speckle reduction method inspired by the recent successes of deep learning in medical imaging. We present two versions of the network to reflect the needs and preferences of different end-users. Specifically, we train a convolution neural network to denoise cross-sections from OCT volumes of healthy eyes using either (1) mean-squared error, or (2) a generative adversarial network (GAN) with Wasserstein distance and perceptual similarity. We then interrogate the success of both methods with extensive quantitative and qualitative metrics on cross-sections from both healthy and glaucomatous eyes. The results show that the former approach provides state-of-the-art improvement in quantitative metrics such as PSNR and SSIM, and aids layer segmentation. However, the latter approach, which puts more weight on visual perception, outperformed for qualitative comparisons based on accuracy, clarity, and personal preference. Overall, our results demonstrate the effectiveness and efficiency of a deep learning approach to denoising OCT images, while maintaining subtle details in the images.

The Henle Fiber Layer in Albinism: Comparison to Normal and Relationship to Outer Nuclear Layer Thickness and Foveal Cone Density.

Purpose: Directional optical coherence tomography (D-OCT) allows the visualization of the Henle fiber layer (HFL) in vivo. Here, we used D-OCT to characterize the HFL and outer nuclear layer (ONL) in albinism and examine the relationship between true foveal ONL and peak cone density. Methods: Horizontal D-OCT B-scans were acquired, registered, and averaged for 12 subjects with oculocutaneous albinism and 26 control subjects. Averaged images were manually segmented to extract HFL and ONL thickness. Adaptive optics scanning light ophthalmoscopy was used to acquire images of the foveal cone mosaic in 10 subjects with albinism, from which peak cone density was assessed. Results: Across the foveal region, the HFL topography was different between subjects with albinism and normal controls. In particular, foveal HFL thickness was thicker in albinism than in normal controls (P < 0.0001), whereas foveal ONL thickness was thinner in albinism than in normal controls (P < 0.0001). The total HFL and ONL thickness was not significantly different between albinism and controls (P = 0.3169). Foveal ONL thickness was positively correlated with peak cone density in subjects with albinism (r = 0.8061, P = 0.0072). Conclusions: Foveal HFL and ONL topography are significantly altered in albinism relative to normal controls. Our data suggest that increased foveal cone packing drives the formation of Henle fibers, more so than the lateral displacement of inner retinal neurons (which is reduced in albinism). The ability to quantify foveal ONL and HFL may help further stratify grading schemes used to assess foveal hypoplasia.

Collaborative SDOCT Segmentation and Analysis Software.

Spectral domain optical coherence tomography (SDOCT) is routinely used in the management and diagnosis of a variety of ocular diseases. This imaging modality also finds widespread use in research, where quantitative measurements obtained from the images are used to track disease progression. In recent years, the number of available scanners and imaging protocols grown and there is a distinct absence of a unified tool that is capable of visualizing, segmenting, and analyzing the data. This is especially noteworthy in longitudinal studies, where data from older scanners and/or protocols may need to be analyzed. Here, we present a graphical user interface (GUI) that allows users to visualize and analyze SDOCT images obtained from two commonly used scanners. The retinal surfaces in the scans can be segmented using a previously described method, and the retinal layer thicknesses can be compared to a normative database. If necessary, the segmented surfaces can also be corrected and the changes applied. The interface also allows users to import and export retinal layer thickness data to an SQL database, thereby allowing for the collation of data from a number of collaborating sites.

Longitudinal Analysis of Mouse SDOCT Volumes.

Spectral-domain optical coherence tomography (SDOCT), in addition to its routine clinical use in the diagnosis of ocular diseases, has begun to find increasing use in animal studies. Animal models are frequently used to study disease mechanisms as well as to test drug efficacy. In particular, SDOCT provides the ability to study animals longitudinally and non-invasively over long periods of time. However, the lack of anatomical landmarks makes the longitudinal scan acquisition prone to inconsistencies in orientation. Here, we propose a method for the automated registration of mouse SDOCT volumes. The method begins by accurately segmenting the blood vessels and the optic nerve head region in the scans using a pixel classification approach. The segmented vessel maps from follow-up scans were registered using an iterative closest point (ICP) algorithm to the baseline scan to allow for the accurate longitudinal tracking of thickness changes. Eighteen SDOCT volumes from a light damage model study were used to train a random forest utilized in the pixel classification step. The area under the curve (AUC) in a leave-one-out study for the retinal blood vessels and the optic nerve head (ONH) was found to be 0.93 and 0.98, respectively. The complete proposed framework, the retinal vasculature segmentation and the ICP registration, was applied to a secondary set of scans obtained from a light damage model. A qualitative assessment of the registration showed no registration failures.

Simultaneous Segmentation of Retinal Surfaces and Microcystic Macular Edema in SDOCT Volumes.

Optical coherence tomography (OCT) is a noninvasive imaging modality that has begun to find widespread use in retinal imaging for the detection of a variety of ocular diseases. In addition to structural changes in the form of altered retinal layer thicknesses, pathological conditions may also cause the formation of edema within the retina. In multiple sclerosis, for instance, the nerve fiber and ganglion cell layers are known to thin. Additionally, the formation of pseudocysts called microcystic macular edema (MME) have also been observed in the eyes of about 5% of MS patients, and its presence has been shown to be correlated with disease severity. Previously, we proposed separate algorithms for the segmentation of retinal layers and MME, but since MME mainly occurs within specific regions of the retina, a simultaneous approach is advantageous. In this work, we propose an automated globally optimal graph-theoretic approach that simultaneously segments the retinal layers and the MME in volumetric OCT scans. SD-OCT scans from one eye of 12 MS patients with known MME and 8 healthy controls were acquired and the pseudocysts manually traced. The overall precision and recall of the pseudocyst detection was found to be 86.0% and 79.5%, respectively.

Voxel Based Morphometry in Optical Coherence Tomography: Validation & Core Findings.

Optical coherence tomography (OCT) of the human retina is now becoming established as an important modality for the detection and tracking of various ocular diseases. Voxel based morphometry (VBM) is a long standing neuroimaging analysis technique that allows for the exploration of the regional differences in the brain. There has been limited work done in developing registration based methods for OCT, which has hampered the advancement of VBM analyses in OCT based population studies. Following on from our recent development of an OCT registration method, we explore the potential benefits of VBM analysis in cohorts of healthy controls (HCs) and multiple sclerosis (MS) patients. Specifically, we validate the stability of VBM analysis in two pools of HCs showing no significant difference between the two populations. Additionally, we also present a retrospective study of age and sex matched HCs and relapsing remitting MS patients, demonstrating results consistent with the reported literature while providing insight into the retinal changes associated with this MS subtype.

Characterizing the Impact of Off-Axis Scan Acquisition on the Reproducibility of Total Retinal Thickness Measurements in SDOCT Volumes.

PURPOSE: Off-axis acquisition of spectral domain optical coherence tomography (SDOCT) images has been shown to increase total retinal thickness (TRT) measurements. We analyzed the reproducibility of TRT measurements obtained using either the retinal pigment epithelium (RPE) or Bruch's membrane as reference surfaces in off-axis scans intentionally acquired through multiple pupil positions. METHODS: Five volumetric SDOCT scans of the macula were obtained from one eye of 25 normal subjects. One scan was acquired through a central pupil position, while subsequent scans were acquired through four peripheral pupil positions. The internal limiting membrane, the RPE, and Bruch's membrane were segmented using automated approaches. These volumes were registered to each other and the TRT was evaluated in 9 Early Treatment of Diabetic Retinopathy Study (ETDRS) regions. The reproducibility of the TRT obtained using the RPE was computed using the mean difference, coefficient of variation (CV), and the intraclass correlation coefficient (ICC), and compared to those obtained using Bruch's membrane as the reference surface. A secondary set of 1545 SDOCT scans was also analyzed in order to gauge the incidence of off-axis scans in a typical acquisition environment. RESULTS: The photoreceptor tips were dimmer in off-axis images, which affected the RPE segmentation. The overall mean TRT difference and CV obtained using the RPE were 7.04 ± 4.31 µm and 1.46%, respectively, whereas Bruch's membrane was 1.16 ± 1.00 µm and 0.32%, respectively. The ICCs at the subfoveal TRT were 0.982 and 0.999, respectively. Forty-one percent of the scans in the secondary set showed large tilts (> 6%). CONCLUSIONS: RPE segmentation is confounded by its proximity to the interdigitation zone, a structure strongly affected by the optical Stiles-Crawford effect. Bruch's membrane, however, is unaffected leading to a more robust segmentation that is less dependent upon pupil position. TRANSLATIONAL RELEVANCE: The way in which OCT images are acquired can independently affect the accuracy of automated retinal thickness measurements. Assessment of scan angle in a clinical dataset demonstrates that off-axis scans are common, which emphasizes the need for caution when relying on automated thickness parameters when this component of scan acquisition is not controlled for.

Automated 3D segmentation of multiple surfaces with a shared hole: segmentation of the neural canal opening in SD-OCT volumes.

The need to segment multiple interacting surfaces is a common problem in medical imaging and it is often assumed that such surfaces are continuous within the confines of the region of interest. However, in some application areas, the surfaces of interest may contain a shared hole in which the surfaces no longer exist and the exact location of the hole boundary is not known a priori. The boundary of the neural canal opening seen in spectral-domain optical coherence tomography volumes is an example of a "hole" embedded with multiple surrounding surfaces. Segmentation approaches that rely on finding the surfaces alone are prone to failures as deeper structures within the hole can "attract" the surfaces and pull them away from their correct location at the hole boundary. With this application area in mind, we present a graph-theoretic approach for segmenting multiple surfaces with a shared hole. The overall cost function that is optimized consists of both the costs of the surfaces outside the hole and the cost of boundary of the hole itself. The constraints utilized were appropriately adapted in order to ensure the smoothness of the hole boundary in addition to ensuring the smoothness of the non-overlapping surfaces. By using this approach, a significant improvement was observed over a more traditional two-pass approach in which the surfaces are segmented first (assuming the presence of no hole) followed by segmenting the neural canal opening.

Automated 3D Segmentation of Intraretinal Surfaces in SD-OCT Volumes in Normal and Diabetic Mice.

PURPOSE: To describe an adaptation of an existing graph-theoretic method (initially developed for human optical coherence tomography [OCT] images) for the three-dimensional (3D) automated segmentation of 10 intraretinal surfaces in mice scans, and assess the accuracy of the method and the reproducibility of thickness measurements. METHODS: Ten intraretinal surfaces were segmented in repeat spectral domain (SD)-OCT volumetric images acquired from normal (n = 8) and diabetic (n = 10) mice. The accuracy of the method was assessed by computing the border position errors of the automated segmentation with respect to manual tracings obtained from two experts. The reproducibility was statistically assessed for four retinal layers within eight predefined regions using the mean and SD of the differences in retinal thickness measured in the repeat scans, the coefficient of variation (CV) and the intraclass correlation coefficients (ICC; with 95% confidence intervals [CIs]). RESULTS: The overall mean unsigned border position error for the 10 surfaces computed over 97 B-scans (10 scans, 10 normal mice) was 3.16 ± 0.91 µm. The overall mean differences in retinal thicknesses computed from the normal and diabetic mice were 1.86 ± 0.95 and 2.15 ± 0.86 µm, respectively. The CV of the retinal thicknesses for all the measured layers ranged from 1.04% to 5%. The ICCs for the total retinal thickness in the normal and diabetic mice were 0.78 [0.10, 0.92] and 0.83 [0.31, 0.96], respectively. CONCLUSION: The presented method (publicly available as part of the Iowa Reference Algorithms) has acceptable accuracy and reproducibility and is expected to be useful in the quantitative study of intraretinal layers in mice. TRANSLATIONAL RELEVANCE: The presented method, initially developed for human OCT, has been adapted for mice, with the potential to be adapted for other animals as well. Quantitative in vivo assessment of the retina in mice allows changes to be measured longitudinally, decreasing the need for them.

A combined machine-learning and graph-based framework for the segmentation of retinal surfaces in SD-OCT volumes.

Optical coherence tomography is routinely used clinically for the detection and management of ocular diseases as well as in research where the studies may involve animals. This routine use requires that the developed automated segmentation methods not only be accurate and reliable, but also be adaptable to meet new requirements. We have previously proposed the use of a graph-theoretic approach for the automated 3-D segmentation of multiple retinal surfaces in volumetric human SD-OCT scans. The method ensures the global optimality of the set of surfaces with respect to a cost function. Cost functions have thus far been typically designed by hand by domain experts. This difficult and time-consuming task significantly impacts the adaptability of these methods to new models. Here, we describe a framework for the automated machine-learning based design of the cost function utilized by this graph-theoretic method. The impact of the learned components on the final segmentation accuracy are statistically assessed in order to tailor the method to specific applications. This adaptability is demonstrated by utilizing the method to segment seven, ten and five retinal surfaces from SD-OCT scans obtained from humans, mice and canines, respectively. The overall unsigned border position errors observed when using the recommended configuration of the graph-theoretic method was 6.45 ± 1.87 µm, 3.35 ± 0.62 µm and 9.75 ± 3.18 µm for the human, mouse and canine set of images, respectively.