Browsing by Subject "Ophthalmology"

Diagnosis, prognosis, and treatment of many ocular and neurodegenerative diseases, including achromatopsia (ACHM), require the visualization of microscopic structures in the eye. The development of adaptive optics ophthalmic imaging systems has made high resolution visualization of ocular microstructures possible. These systems include the confocal and split detector adaptive optics scanning light ophthalmoscope (AOSLO), which can visualize human cone and rod photoreceptors in vivo. However, the avalanche of data generated by such imaging systems is often too large, costly, and time consuming to be evaluated manually, making automation necessary. The few currently available automated cone photoreceptor identification methods are unable to reliably identify rods and cones in low-quality images of diseased eyes, which are common in clinical practice.

This dissertation describes the development of automated methods for the analysis of AOSLO images, specifically focusing on cone and rod photoreceptors which are the most commonly studied biomarker using these systems. A traditional image processing approach, which requires little training data and takes advantage of intuitive image features, is presented for detecting cone photoreceptors in split detector AOSLO images. The focus is then shifted to deep learning using convolutional neural networks (CNNs), which have been shown in other image processing tasks to be more adaptable and produce better results than classical image processing approaches, at the cost of requiring more training data and acting as a “black box”. A CNN based method for detecting cones is presented and validated against state-of-the-art cone detections methods for confocal and split detector images. The CNN based method is then modified to take advantage of multimodal AOSLO information in order to detect cones in images of subjects with ACHM. Finally, a significantly faster CNN based approach is developed for the classification and detection of cones and rods, and is validated on images from both healthy and pathological subjects. Additionally, several image processing and analysis works on optical coherence tomography images that were carried out during the completion of this dissertation are presented.

The completion of this dissertation led to fast and accurate image analysis tools for the quantification of biomarkers in AOSLO images pertinent to an array of retinal diseases, lessening the reliance on subjective and time-consuming manual analysis. For the first time, automatic methods have comparable accuracy to humans for quantifying photoreceptors in diseased eyes. This is an important step in the long-term goal to facilitate early diagnosis, accurate prognosis, and personalized treatment of ocular and neurodegenerative diseases through optimal visualization and quantification of microscopic structures in the eye.

Age-related macular degeneration (AMD), a major cause of irreversible vision loss in elderly populations, has been associated genetically with the complement system, and especially with complement factor H (CFH). However, it remains unknown how CFH affects risk of developing AMD. Humans additionally express a truncated splice variant of CFH, Factor H-like 1 (FHL-1). The retinal pigmented epithelium (RPE) is separated from circulation by Bruch’s membrane (BrM), and it has been reported that CFH is unable to diffuse through BrM, while FHL-1 can. Therefore, it is uncertain whether FHL-1 is responsible for most complement regulation in BrM and on RPE cells, or if local production of CFH is more important. We examined this question in Cfh-/- animals through the use of AAVs, allowing expression of both FHL-1 or truncated versions of CFH retaining the C-terminal end of protein, 18tCFH and 12tCFH, from either local sources following subretinal injection or from systemic sources following tail vein injection. While all constructs showed regulation of the complement system using Western blots in circulation, only 18tCFH, the most complete of the constructs, showed definitive regulation of the complement system following subretinal injection. Additionally, there was more evidence of complement regulation following expression locally in the eye, as opposed to circulation. These results suggest that the C-terminus of the protein is essential for complement regulation in the eye, and that local delivery of the protein is the better delivery route. These findings are important for the design of future gene therapies for complement regulation in AMD patients.

Background: Because most developing countries lack sufficient resources and infrastructure to conduct population-based studies on childhood blindness, it can be difficult to obtain epidemiologically reliable data available for planning public health strategies to effectively address the major determinants of childhood blindness. The major etiologies of blindness can differ regionally and intra-regionally. The objective of this retrospective study was to determine (1) the major causes of childhood blindness (BL) and severe visual impairment (SVI) in students who attend Wa Methodist School for the Blind in Upper West Region, North Ghana, and (2) any potential temporal trends in the causes of blindness for this region.

Methods: In this retrospective study, demographic data and clinical information from an eye screening at Wa Methodist School for the Blind were coded according to the World Health Organization/Prevention of Blindness standardized reporting methodology. Causes of BL and SVI were categorized anatomically and etiologically. We determined the major causes of BL/SVI over time using information provided about the age at onset of visual loss for each student.

Results: The major anatomical causes of BL/SVI among the 190 students screened were corneal opacity and phthisis bulbi (n=28, 15%), optic atrophy (n=23, 13%), glaucoma (n=18, 9%), microphthalmos (n=18, 9%), and cataract (n=18, 9%). Within the first year of life, students became blind mainly due to whole globe causes (n=23, 26%), cataract (n=15, 17%), and optic atrophy (n=11, 13%). Those who became blind after age one year had whole globe causes (n=26, 26%), corneal opacity (n=24, 24%), and optic atrophy (n=13, 13%).

Conclusion: At the Wa Methodist School for the Blind, the major anatomical causes of BL/SVI were corneal opacity and phthisis bulbi. About half of all students became blind within the first year of life, and were disproportionately affected by cataract and retinal causes in comparison to the other students who became blind after age one year. While research in blind schools has a number of implicit disadvantages and limitations, considering the temporal trends and other epidemiological factors of blindness may increase the usefulness and/or implications of the data that come from blind school studies in order to improve screening methods for newborns in hospitals and primary care centers, and to help tailor preventative and treatment programs to reduce avoidable childhood blindness in neonates and schoolchildren.

Retinal and corneal diseases are the leading causes of global vision impairment. The advent of high‐resolution ophthalmic imaging technology enables the visualization of internal and external structures of the eye, thereby aiding clinicians in the assessment of these diseases. However, there are high costs and constraints associated with manual image analysis by clinicians. Therefore, the development of automatic algorithms with suitable performance for clinical practice is crucial to alleviate the burden on clinicians and improve the efficiency of the clinical workflow. This dissertation describes the development of a deep learning‐based image analysis framework consisting of computational algorithms for accurate automatic medical image analysis applied to ophthalmology for the clinical assessment of retinal and corneal diseases.

In Chapter 2, we developed longitudinal algorithms for the assessment of 3‐D medical images with applications for optical coherence tomography (OCT). For clinical application, we used the algorithms for the assessment of macular telangiectasia type 2 and USH2A‐related retinitis pigmentosa in large‐scale clinical trials. We also introduced the concept of longitudinal transfer learning to develop personalized algorithms and introduced a new paradigm for validating automatic algorithms for clinical applications beyond performance metrics.

In Chapter 3, we developed region‐based algorithms for the assessment of 2‐D medical images with applications for slit lamp photography (SLP). For clinical application, we used the algorithms for the assessment of microbial keratitis in clinical studies from the USA and India. We also demonstrated the potential of automatic SLP-based measurements for assessing ocular function and paved the way for the development of objective and standardized strategies for the assessment of corneal diseases.

In Chapter 4, we developed hybrid algorithms for the joint assessment of spatially-registered 2‐D and 3‐D medical images with applications for indocyanine green angiography and OCT. For clinical application, we used the algorithms for the assessment of polypoidal choroidal vasculopathy, a sub‐type of age‐related macular degeneration. We introduced hybrid network architectures with fusion attention modules that effectively processed co‐registered images of different dimensionalities to enable sharing of learned features between the different imaging modalities. We also derived quantitative definitions of important imaging biomarkers of the disease.

In conclusion, this dissertation provides an image analysis framework for clinical management of retinal and corneal diseases. Our deep learning‐based computational algorithms can accurately identify and quantify important disease biomarkers automatically and have been validated for several aspects of clinical applicability. These algorithms can be used by clinicians to improve the efficiency of the clinical workflow, leading to timely and precise medical decisions, ultimately improving patient outcomes.

Optical coherence tomography (OCT) is a non-invasive optical imaging modality that provides micron-scale resolution of tissue micro-structure over depth ranges of several millimeters. This imaging technique has had a profound effect on the field of ophthalmology, wherein it has become the standard of care for the diagnosis of many retinal pathologies. Applications of OCT in the anterior eye, as well as for imaging of coronary arteries and the gastro-intestinal tract, have also shown promise, but have not yet achieved widespread clinical use.

The usable imaging depth of OCT systems is most often limited by one of three factors: optical attenuation, inherent imaging range, or depth-of-focus. The first of these, optical attenuation, stems from the limitation that OCT only detects singly-scattered light. Thus, beyond a certain penetration depth into turbid media, essentially all of the incident light will have been multiply scattered, and can no longer be used for OCT imaging. For many applications (especially retinal imaging), optical attenuation is the most restrictive of the three imaging depth limitations. However, for some applications, especially anterior segment, cardiovascular (catheter-based) and GI (endoscopic) imaging, the usable imaging depth is often not limited by optical attenuation, but rather by the inherent imaging depth of the OCT systems. This inherent imaging depth, which is specific to only Fourier Domain OCT, arises due to two factors: sensitivity fall-off and the complex conjugate ambiguity. Finally, due to the trade-off between lateral resolution and axial depth-of-focus inherent in diffractive optical systems, additional depth limitations sometimes arises in either high lateral resolution or extended depth OCT imaging systems. The depth-of-focus limitation is most apparent in applications such as adaptive optics (AO-) OCT imaging of the retina, and extended depth imaging of the ocular anterior segment.

In this dissertation, techniques for extending the imaging range of OCT systems are developed. These techniques include the use of a high spectral purity swept source laser in a full-field OCT system, as well as the use of a peculiar phenomenon known as coherence revival to resolve the complex conjugate ambiguity in swept source OCT. In addition, a technique for extending the depth of focus of OCT systems by using a polarization-encoded, dual-focus sample arm is demonstrated. Along the way, other related advances are also presented, including the development of techniques to reduce crosstalk and speckle artifacts in full-field OCT, and the use of fast optical switches to increase the imaging speed of certain low-duty cycle swept source OCT systems. Finally, the clinical utility of these techniques is demonstrated by combining them to demonstrate high-speed, high resolution, extended-depth imaging of both the anterior and posterior eye simultaneously and in vivo.

Alzheimer’s disease (AD) is a neurodegenerative disease currently affecting 5.8 million Americans and more than 50 million people worldwide. It is a progressive disease that destroys cognitive functions, leading to dementia. With increasing life-expectancy, important efforts have been made to clinically diagnose this age-related disease. However, definitive diagnosis of AD has been challenging, especially at an early stage, as there is a lack of quantifiable changes. Recently, many researchers have shown retinal changes as an extension of the brain pathology, leading to a window to study AD using fast and high-resolution retinal imaging tools. This dissertation will be focused on the development of low-cost imaging tools aimed to extract retinal biomarkers for AD. Specifically, the use of optical coherence tomography (OCT) and angle-resolved low-coherence interferometry (a/LCI) will be described, with steps leading to a combined optical system for retinal imaging in humans. OCT has already been established as the gold standard in ophthalmology due to its excellent axial resolution and high sensitivity. Similar to OCT, a/LCI is another interferometric technique that provides depth resolution. Previous work has supported the ability of a/LCI to retrieve depth-resolved light scattering measurements of nuclear morphology in dysplastic tissue. The use of OCT as image guidance for a/LCI can strengthen the technique, providing sample orientation as well as retinal layer segmentation to pinpoint a/LCI measurements. The dissertation starts with the development and clinical application of a low-cost OCT system. Despite the prevalence of OCT, its high-cost nature has limited its access to large eye centers and away from low-resource settings. Clinical feasibility of a complete low-cost OCT system will be evaluated, and its imaging performance compared to a commercial system. System design will be discussed, followed by a comprehensive image processing pipeline to characterize image quality for subsequent low-cost systems. The subsequent portions outline studies using a/LCI and the extraction of light scattering parameters in an AD mouse model. A benchtop co-registered system using a/LCI guided by OCT allowed measurements of depth-resolved light scattering measurements in an AD mouse retina model. Resulting parameters serve as unique quantification of AD tissue structure with potential to be translated to future human studies. A scanning mechanism for 2D a/LCI is also presented, which also allowed for the characterization of a/LCI sensitivity to anisotropic scattering that is often present in the complex retinal tissue. The last portion discusses the development of a second-generation low-cost OCT system which will be integrated in a combined imaging system for eventual AD studies in human patients. Several technical improvements are shown to facilitate clinical retinal imaging at the point-of-care. A characterization of this system in a small clinical study will illustrate the system’s capability to screen AD patients, and to serve as a morphological image guide for a clinical a/LCI system. Finally, a discussion of how the low-cost OCT system can be integrated to a multimodal imaging system for AD human retinal biomarker extraction will be provided.

Advancements in retinal imaging technologies have drastically improved the quality of eye care in the past couple decades. Scanning laser ophthalmoscopy (SLO) and optical coherence tomography (OCT) are two examples of critical imaging modalities for the diagnosis of retinal pathologies. However current-generation SLO and OCT systems have limitations in diagnostic capability due to the following factors: the use of bulky tabletop systems, monochromatic imaging, and resolution degradation due to ocular aberrations and diffraction.

Bulky tabletop SLO and OCT systems are incapable of imaging patients that are supine, under anesthesia, or otherwise unable to maintain the required posture and fixation. Monochromatic SLO and OCT imaging prevents the identification of various color-specific diagnostic markers visible with color fundus photography like those of neovascular age-related macular degeneration. Resolution degradation due to ocular aberrations and diffraction has prevented the imaging of photoreceptors close to the fovea without the use of adaptive optics (AO), which require bulky and expensive components that limit the potential for widespread clinical use.

In this dissertation, techniques for extending the diagnostic capability of SLO and OCT systems are developed. These techniques include design strategies for miniaturizing and combining SLO and OCT to permit multi-modal, lightweight handheld probes to extend high quality retinal imaging to pediatric eye care. In addition, a method for extending true color retinal imaging to SLO to enable high-contrast, depth-resolved, high-fidelity color fundus imaging is demonstrated using a supercontinuum light source. Finally, the development and combination of SLO with a super-resolution confocal microscopy technique known as optical photon reassignment (OPRA) is demonstrated to enable high-resolution imaging of retinal photoreceptors without the use of adaptive optics.

Fuchs endothelial corneal dystrophy (FECD) is a complex, late-onset disorder that is a frequent indication for corneal transplantation and affects women more frequently than men. Although linkage and association studies in patients of European and Asian ancestry have started to explain the genetic basis of this disorder, the mechanism by which FECD develops is still unclear. Three projects were undertaken to help elucidate the genetic basis of FECD. The first project examined a large, multigenerational family that exhibited strong familial clustering of FECD and identified evidence of linkage to chromosome 18. This locus that has also been implicated through work on large and small FECD families as well as unrelated patients. The second project examined African-Americans with FECD and is the first work to examine this population of patients with respect to the FECD phenotype. Novel variants in three FECD candidate genes, COL8A2, SLC4A11, and ZEB1 were identified at approximately the same rate as observed in patients of other ancestries, reinforcing the notion that these genes only contribute to a small fraction of FECD genetic susceptibility. Finally, the influence of environmental factors on FECD susceptibility was examined through the use of a risk factor questionnaire given to cases and controls at the time of study enrollment. Several factors, including gender, age, and cataracts, were found to significantly affect FECD risk. Gender and the number of cataract surgeries were found to significantly interact with a genetic variant, rs613872 in the TCF4 locus on chromosome 18 that has been consistently and reproducibly associated with FECD, to influence FECD susceptibility. Together, these findings indicate that the genetic basis of FECD is complex, and recent advances in the field show promise in accelerating the pace of discovery that will hopefully develop better FECD treatments in the near future.

Optical coherence tomography (OCT) is a non-invasive imaging modality that uses low coherence interferometry to generate three-dimensional datasets of a sample's structure. OCT has found tremendous clinical applications in imaging the retina and has demonstrated great utility in the diagnosis of various retinal diseases. However, such diagnoses rely upon the ability to observe abnormalities in the structure of the retina caused by pathology. By the time an ocular disease has progressed to the point of affecting the morphology of the retina, irreversible vision loss in the eye may already occur. Changes in the functionality of the tissue often precede changes to the structure. Thus, if imaging methods are developed to provide additional functional information about the behavior and response of the retinal tissue and vasculature, earlier treatment for disease may be prescribed, thus preserving vision for the patient.

Within the last decade, significant technological advances in OCT systems have enabled high-speed and high sensitivity image acquisition using either spectral domain OCT (SDOCT) or swept-source OCT (SSOCT) configurations. Such systems use Fourier processing to extract structural information of a sample from interferometric principles. But such systems also have access to the optical phase information, which allows for functional analysis of sample dynamics. This dissertation details the development and application of methods using both intensity and phase information as a tool for studying interesting biological phenomena. The goal of this work is an extension of techniques to image the vasculature in the retina and enhance the clinical utility of OCT.

I first outline basic theory necessary for understanding the principles of OCT. I then describe OCT phase imaging in cellular applications as a demonstration of the ability of OCT to provide functional information on biological dynamics. Phase imaging methods suffer from an artifact known as phase wrapping, and I have developed a software technique to overcome this problem in OCT, thus extending its usefulness in providing quantitative information. I characterize the limitations in measuring moving scatterers with Doppler OCT in both SDOCT and SSOCT system. I also show the ability to image the vasculature in the retina using variance imaging with a high-speed retinal imaging system and software based methods to correct for patient motion and create a widefield mosaic in an automated manner. Finally, future directions for this work are discussed.

The goal of my Ph.D. thesis is to enhance the visualization of the peripheral retina using wide-field optical coherence tomography (OCT) in a clinical setting.

OCT has gain widespread adoption in clinical ophthalmology due to its ability to visualize the diseases of the macula and central retina in three-dimensions, however, clinical OCT has a limited field-of-view of 300. There has been increasing interest to obtain high-resolution images outside of this narrow field-of-view, because three-dimensional imaging of the peripheral retina may prove to be important in the early detection of neurodegenerative diseases, such as Alzheimer's and dementia, and the monitoring of known ocular diseases, such as diabetic retinopathy, retinal vein occlusions, and choroid masses.

Before attempting to build a wide-field OCT system, we need to better understand the peripheral optics of the human eye. Shack-Hartmann wavefront sensors are commonly used tools for measuring the optical imperfections of the eye, but their acquisition speed is limited by their underlying camera hardware. The first aim of my thesis research is to create a fast method of ocular wavefront sensing such that we can measure the wavefront aberrations at numerous points across a wide visual field. In order to address aim one, we will develop a sparse Zernike reconstruction technique (SPARZER) that will enable Shack-Hartmann wavefront sensors to use as little as 1/10th of the data that would normally be required for an accurate wavefront reading. If less data needs to be acquired, then we can increase the speed at which wavefronts can be recorded.

For my second aim, we will create a sophisticated optical model that reproduces the measured aberrations of the human eye. If we know how the average eye's optics distort light, then we can engineer ophthalmic imaging systems that preemptively cancel inherent ocular aberrations. This invention will help the retinal imaging community to design systems that are capable of acquiring high resolution images across a wide visual field. The proposed model eye is also of interest to the field of vision science as it aids in the study of how anatomy affects visual performance in the peripheral retina.

Using the optical model from aim two, we will design and reduce to practice a clinical OCT system that is capable of imaging a large (800) field-of-view with enhanced visualization of the peripheral retina. A key aspect of this third and final aim is to make the imaging system compatible with standard clinical practices. To this end, we will incorporate sensorless adaptive optics in order to correct the inter- and intra- patient variability in ophthalmic aberrations. Sensorless adaptive optics will improve both the brightness (signal) and clarity (resolution) of features in the peripheral retina without affecting the size of the imaging system.

The proposed work should not only be a noteworthy contribution to the ophthalmic and engineering communities, but it should strengthen our existing collaborations with the Duke Eye Center by advancing their capability to diagnose pathologies of the peripheral retinal.

Accurate quantification of anatomical and pathological structures in the eye is crucial for the study and diagnosis of potentially blinding diseases. Earlier and faster detection of ophthalmic imaging biomarkers also leads to optimal treatment and improved vision recovery. While modern optical imaging technologies such as optical coherence tomography (OCT) and adaptive optics (AO) have facilitated in vivo visualization of the eye at the cellular scale, the massive influx of data generated by these systems is often too large to be fully analyzed by ophthalmic experts without extensive time or resources. Furthermore, manual evaluation of images is inherently subjective and prone to human error.

This dissertation describes the development and validation of a framework called graph theory and dynamic programming (GTDP) to automatically detect and quantify ophthalmic imaging biomarkers. The GTDP framework was validated as an accurate technique for segmenting retinal layers on OCT images. The framework was then extended through the development of the quasi-polar transform to segment closed-contour structures including photoreceptors on AO scanning laser ophthalmoscopy images and retinal pigment epithelial cells on confocal microscopy images.

The GTDP framework was next applied in a clinical setting with pathologic images that are often lower in quality. Algorithms were developed to delineate morphological structures on OCT indicative of diseases such as age-related macular degeneration (AMD) and diabetic macular edema (DME). The AMD algorithm was shown to be robust to poor image quality and was capable of segmenting both drusen and geographic atrophy. To account for the complex manifestations of DME, a novel kernel regression-based classification framework was developed to identify retinal layers and fluid-filled regions as a guide for GTDP segmentation.

The development of fast and accurate segmentation algorithms based on the GTDP framework has significantly reduced the time and resources necessary to conduct large-scale, multi-center clinical trials. This is one step closer towards the long-term goal of improving vision outcomes for ocular disease patients through personalized therapy.

The human retina is essential to quality of life and therefore a topic of intense clinical and research interest. The combination of this interest with modern biophotonics has yielded a number of technological and medical developments now in various stages of adoption.

Optical coherence tomography (OCT) is a noninvasive optical imaging technique that utilizes coherent light to produce 3-D images with resolutions as fine as a micrometer. Since its invention in 1990, it has become part of the standard of care in opthalmology, shedding new light on the progression of diseases, therapeutic efficacy, childhood development, and real-time surgery in the retina. OCT has also found applications in microscopy, cardiology, pulmonology, and many other fields.

OCT has become valuable for the standard of care primarily due to its abilities to visualize the structural and functional layers of the retina. The thicknesses and volumes of certain can be used as diagnostic criteria and thus there is a high demand of OCT image assessment. In response, many researchers have developed software algorithms to automatically identify and mark, or segment, each layer.

Scanning light ophthalmoscopy or scanning laser ophthalmoscopy (SLO) is similar to OCT but uses confocal gating to produce high-contrast high-speed en face images of the retina. Although SLO has not become as prevalent as OCT in the clinic, it is frequently combined with adaptive optics (AO) to produce extremely high-resolution images of rod and cone photoreceptors, ganglion cells, and moving blood cells in the living retina.

AO is a technique to eliminate image blurring due to monochromatic aberrations in optical systems. By using a spatial light modulator, such as a deformable mirror or liquid crystal array, the wavefront of a beam sent into the eye can be engineered to compensate for the eye's aberrations. AO-SLO was initially developed in 2002 and has continued to be a field of research growth and interest. However, the majority of AOSLO systems require a dedicated room and staff, hindering their clinical adoption

The objective of the work presented herein was to explore the limits of the above imaging modalities. First, we explored the limits of OCT segmentation and demonstrated that the field of automated segmentation is far from its accuracy limit. Second, we explored the limits on SLO portability and developed both the world's smallest SLO probe and the first handheld AOSLO probe. Finally, we explored the limits of SLO resolution, developing the first super-resolution human retinal imaging system through the use of optical reassignment (OR) SLO.

With the advent of Machine Learning and the existence of Electronic Health Records, with most non-federal acute care hospitals a large number of office-based physicians already having opted for having a certified EHRs, each patient has essentially become a big data problem for medical predictions. This is also true in the field of Ophthalmology with its various specific modalities. Across two projects we explore how electronic health records can be used to make predictive model for various condition using machine learning.

Using patient histories and demographics such as age, gender, and race, body mass index (BMI), medications, biologicals, comorbidities, past medical history, and visual acuities we model a risk classifier for progression of age-related macular degeneration from its dry for to its wet form, which is a much faster progressing form of the disease. We found older age, use of biologicals such as anti-VEGF agents, and lover visual acuity to be associated with increased risk of progression of the disease. Our model gave an indicative tool with accuracy of 0.778±0.045, F1 score of 0.795±0.038 and sensitivity of 0.86±0.068. Also using imaging modalities such as SD-OCTs we model the detection of hydro-chloroquine toxicity related retinopathy, and attempt propose a prediction model. Our Model was able to detect hydro-chloroquine toxicity induced retinopathy with a precision of 0.72, recall of 0.92, F1 score of 0.81, and accuracy of 0.81.

Using the two projects we showed that using data extracted from electronic health records we can make effective models for various tasks using machine learning fairly well.