Nonlinear distortion correction for posterior eye segment optical coherence tomography with application to tree shrews

Image segmentation of mouse eye in vivo with optical coherence tomography based on Bayesian classification

OBJECTIVES

Optical coherence tomography (OCT) is a new imaging technology that uses an optical analog of ultrasound imaging for biological tissues. Image segmentation plays an important role in dealing with quantitative analysis of medical images.

METHODS

We have proposed a novel framework to deal with the low intensity problem, based on the labeled patches and Bayesian classification (LPBC) model. The proposed method includes training and testing phases. During the training phase, firstly, we manually select the sub-images of background and Region of Interest (ROI) from the training image, and then extract features by patches. Finally, we train the Bayesian model with the features. The segmentation threshold of each patch is computed by the learned Bayesian model.

RESULTS

In addition, we have collected a new dataset of mouse eyes in vivo with OCT, named MEVOCT, which can be found at URL https://17861318579.github.io/LPBC. MEVOCT consists of 20 high-resolution images. The resolution of every image is 2048 × 2048 pixels.

CONCLUSIONS

The experimental results demonstrate the effectiveness of the LPBC method on the new MEVOCT dataset. The ROI segmentation is of great importance for the distortion correction.

Heterogenous thinning of peripapillary tissues occurs early during high myopia development in juvenile tree shrews

Myopia is an independent risk factor for glaucoma, but the link between both conditions remains unknown. Both conditions induce connective tissue remodeling at the optic nerve head (ONH), including the peripapillary tissues. The purpose of this study was to investigate the thickness changes of the peripapillary tissues during experimental high myopia development in juvenile tree shrews. Six juvenile tree shrews experienced binocular normal vision, while nine received monocular -10D lens treatment starting at 24 days of visual experience (DVE) to induce high myopia in one eye and the other eye served as control. Daily refractive and biometric measurements and weekly optical coherence tomography scans of the ONH were obtained for five weeks. Peripapillary sclera (Scl), choroid-retinal pigment epithelium complex (Ch-RPE), retinal nerve fiber layer (RNFL), and remaining retinal layers (RRL) were auto-segmented using a deep learning algorithm after nonlinear distortion correction. Peripapillary thickness values were quantified from 3D reconstructed segmentations. All lens-treated eyes developed high myopia (-9.8 ± 1.5 D), significantly different (P < 0.001) from normal (0.69 ± 0.45 D) and control eyes (0.76 ± 1.44 D). Myopic eyes showed significant thinning of all peripapillary tissues compared to both, normal and control eyes (P < 0.001). At the experimental end point, the relative thinning from baseline was heterogeneous across tissues and significantly more pronounced in the Scl (-8.95 ± 3.1%) and Ch-RPE (-16.8 ± 5.8%) when compared to the RNFL (-5.5 ± 1.6%) and RRL (-6.7 ± 1.8%). Furthermore, while axial length increased significantly throughout the five weeks of lens wear, significant peripapillary tissue thinning occurred only during the first week of the experiment (until a refraction of -2.5 ± 1.9 D was reached) and ceased thereafter. A sectorial analysis revealed no clear pattern. In conclusion, our data show that in juvenile tree shrews, experimental high myopia induces significant and heterogeneous thinning of the peripapillary tissues, where the retina seems to be protected from profound thickness changes as seen in Ch-RPE and Scl. Peripapillary tissue thinning occurs early during high myopia development despite continued progression of axial elongation. The observed heterogeneous thinning may contribute to the increased risk for pathological optic nerve head remodeling and glaucoma later in life.

Are Macula or Optic Nerve Head Structures Better at Diagnosing Glaucoma? An Answer Using Artificial Intelligence and Wide-Field Optical Coherence Tomography

Purpose

We wanted to develop a deep-learning algorithm to automatically segment optic nerve head (ONH) and macula structures in three-dimensional (3D) wide-field optical coherence tomography (OCT) scans and to assess whether 3D ONH or macula structures (or a combination of both) provide the best diagnostic power for glaucoma.

Methods

A cross-sectional comparative study was performed using 319 OCT scans of glaucoma eyes and 298 scans of nonglaucoma eyes. Scans were compensated to improve deep-tissue visibility. We developed a deep-learning algorithm to automatically label major tissue structures, trained with 270 manually annotated B-scans. The performance was assessed using the Dice coefficient (DC). A glaucoma classification algorithm (3D-CNN) was then designed using 500 OCT volumes and corresponding automatically segmented labels. This algorithm was trained and tested on three datasets: cropped scans of macular tissues, those of ONH tissues, and wide-field scans. The classification performance for each dataset was reported using the area under the curve (AUC).

Results

Our segmentation algorithm achieved a DC of 0.94 ± 0.003. The classification algorithm was best able to diagnose glaucoma using wide-field scans, followed by ONH scans, and finally macula scans, with AUCs of 0.99 ± 0.01, 0.93 ± 0.06 and 0.91 ± 0.11, respectively.

Conclusions

This study showed that wide-field OCT may allow for significantly improved glaucoma diagnosis over typical OCTs of the ONH or macula.

Translational Relevance

This could lead to mainstream clinical adoption of 3D wide-field OCT scan technology.

Longitudinal Changes of Bruch's Membrane Opening, Anterior Scleral Canal Opening, and Border Tissue in Experimental Juvenile High Myopia

Purpose

To investigate the relative positional changes between the Bruch's membrane opening (BMO) and the anterior scleral canal opening (ASCO), and border tissue configuration changes during experimental high myopia development in juvenile tree shrews.

Methods

Juvenile tree shrews were assigned randomly to two groups: binocular normal vision (n = 9) and monocular -10 D lens treatment starting at 24 days of visual experience to induce high myopia in one eye while the other eye served as control (n = 12). Refractive and biometric measurements were obtained daily, and 48 radial optical coherence tomography B-scans through the center of the optic nerve head were obtained weekly for 6 weeks. ASCO and BMO were segmented manually after nonlinear distortion correction.

Results

Lens-treated eyes developed high degree of axial myopia (-9.76 ± 1.19 D), significantly different (P < 0.001) from normal (0.34 ± 0.97 D) and control eyes (0.39 ± 0.88 D). ASCO-BMO centroid offset gradually increased and became significantly larger in the experimental high myopia group compared with normal and control eyes (P < 0.0001) with an inferonasal directional preference. The border tissue showed a significantly higher tendency of change from internally to externally oblique configuration in the experimental high myopic eyes in four sectors: nasal, inferonasal, inferior, and inferotemporal (P < 0.005).

Conclusions

During experimental high myopia development, progressive relative deformations of ASCO and BMO occur simultaneously with changes in border tissue configuration from internally to externally oblique in sectors that are close to the posterior pole (nasal in tree shrews). These asymmetric changes may contribute to pathologic optic nerve head remodeling and an increased risk of glaucoma later in life.

Reverse translation of artificial intelligence in glaucoma: Connecting basic science with clinical applications

Artificial intelligence (AI) has been approved for biomedical research in diverse areas from bedside clinical studies to benchtop basic scientific research. For ophthalmic research, in particular glaucoma, AI applications are rapidly growing for potential clinical translation given the vast data available and the introduction of federated learning. Conversely, AI for basic science remains limited despite its useful power in providing mechanistic insight. In this perspective, we discuss recent progress, opportunities, and challenges in the application of AI in glaucoma for scientific discoveries. Specifically, we focus on the research paradigm of reverse translation, in which clinical data are first used for patient-centered hypothesis generation followed by transitioning into basic science studies for hypothesis validation. We elaborate on several distinctive areas of research opportunities for reverse translation of AI in glaucoma including disease risk and progression prediction, pathology characterization, and sub-phenotype identification. We conclude with current challenges and future opportunities for AI research in basic science for glaucoma such as inter-species diversity, AI model generalizability and explainability, as well as AI applications using advanced ocular imaging and genomic data.

Correction of refractive distortion in whole-eye optical coherence tomography imaging of the mouse eye

Optical coherence tomography (OCT) is an imaging modality that acquires high-resolution cross-sectional images of living tissues and it has become the standard in ophthalmological diagnoses. However, most quantitative morphological measurements are based on the raw OCT images which are distorted by several mechanisms such as the refraction of probe light in the sample and the scan geometries and thus the analysis of the raw OCT images inevitably induced calculation errors. In this paper, based on Fermat's principle and the concept of inverse light tracing, image distortions due to refraction occurred at tissue boundaries in the whole-eye OCT imaging of mouse by telecentric scanning were corrected. Specially, the mathematical correction models were deducted for each interface, and the high-precision whole-eye image was recovered segment by segment. We conducted phantom and in vivo experiments on mouse and human eyes to verify the distortion correction algorithm, and several parameters of the radius of curvature, thickness of tissues and error, were calculated to quantitatively evaluate the images. Experimental results demonstrated that the method can provide accurate and reliable measurements of whole-eye parameters and thus be a valuable tool for the research and clinical diagnosis.

Cone Photoreceptor Loss in Light-Damaged Albino Rats

We investigated the etiology of decreased cone-driven vision in a light damage (LD) model of retinal degeneration. To induce slow, moderate degeneration, albino rats underwent low-intensity light exposure for 10 days. Electroretinography was utilized to assess physiologic function of the rod- and cone-driven retinal function in LD and control rats. Immunohistochemistry targeting cone arrestin allowed for quantification of cone density and for comparison of the decline in function. Photoreceptor loss was quantified by outer nuclear layer thickness decreases, as observed by optical coherence tomography and histology. The LD rats showed decreased rod- and cone-driven function with partial recovery 30 days after cessation of light exposure. In addition, LD rats showed decreased cone photoreceptor densities in the central retinal region compared to control rats. Our results demonstrate that the loss of cone-driven visual function induced by light damage is at least partially due to the death of cone photoreceptors.

Scleral crosslinking using genipin can compromise retinal structure and function in tree shrews

Scleral crosslinking using genipin has been identified as a promising treatment approach for myopia control. The efficacy of genipin to alter biomechanical properties of the sclera has been shown in several animal models of myopia but its safety profile remains unclear. In this safety study, we aim to investigate the effect of scleral crosslinking using retrobulbar injections of genipin on retinal structure and function at genipin doses that were shown to be effective in slowing myopia progression in juvenile tree shrews. To this end, three or five retrobulbar injections of genipin at 0 mM (sham), 10 mM, or 20 mM were performed in one eye every other day. Form deprivation myopia was induced in the injected eye. We quantified retinal function changes using full-field electroretinography and retinal structure changes using in vivo optical coherence tomography imaging and ex vivo histology. The optical coherence tomography results revealed significant thinning of the peripapillary retinal nerve fiber layer in all genipin treated groups including the lowest dose group, which showed no significant treatment effect in slowing myopia progression. In contrast, inducing form deprivation myopia alone and in combination with sham injections caused no obvious thinning of the retinal nerve fiber layer. Electroretinography results showed a significant desensitizing shift of the b-wave semi-saturation constant in the sham group and the second highest genipin dose group, and a significant reduction in b-wave maxima in the two highest genipin dose groups. The ex vivo histology revealed noticeable degeneration of photoreceptors and retinal pigment epithelium in one of two investigated eyes of the highest genipin dose group. While scleral crosslinking using genipin may still be a feasible treatment option for myopia control, our results suggest that repeated retrobulbar injections of genipin at 10 mM or higher are not safe in tree shrews. An adequate and sustained delivery strategy of genipin at lower concentrations will be needed to achieve a safe and effective scleral crosslinking treatment for myopia control in tree shrews. Caution should be taken if the proposed treatment approach is translated to humans.