Oxyopia Vision Science Seminars

Oxyopia Seminars (2016-2017)

Oxyopia, the Greek word meaning "acute vision," is the title of the weekly vision science seminars presented at the Indiana University School of Optometry in conjunction with graduate course V765. The seminars serve a twofold purpose in that they:
  1. stimulate intellectual activity among the faculty, and
  2. provide a learning environment for graduate students.

Oxyopia presenters are IU School of Optometry faculty members and graduate students as well as visiting lecturers from other departments, universities, research facilities, private practices, industry, etc.



Christine A. Curcio

Visualizing the neurodegeneration of age-related macular degeneration in histology and multimodal imaging The Project MACULA website is an online digital microscope of 142 maculas from human donor eyes, 82 with age-related macular degeneration (AMD), a major cause of vision loss in older adults. Systematic and unbiased review of these eyes enabled a first-ever high-resolution histological description of the neurodegeneration of advanced AMD. I will review recent and forthcoming publications that describe for the first time the fine structure of photoreceptor and retinal pigment epithelium (RPE) degeneration in AMD.  Data allow us to introduce RPE transdifferentiation to a migratory phenotype as an important precursor to geographic atrophy. These phenomena are visible in AMD patients using optical coherence tomography, as informed by multimodal imaging.
9/9/2016 Patrice Tankam Endothelial Corneal Dystrophy: The perspective of Regenerative Medicine

Corneal endothelium represents a key component of our visual system in maintaining the corneal transparency. Loss of endothelial function leads to greater hydration of the corneal stroma, which can cause corneal edema, opacity, and blindness. Fuchs’ Endothelial Corneal Dystrophy (FECD) is the most common cause of endothelial dysfunction, which is diagnosed by the appearance of drops (guttae) on the posterior surface of the Descemet’s membrane, yielding a decrease in corneal endothelial cell density. Corneal transplantation is nowadays the mainstay of treatment for endothelial cell dysfunction. The reason is that, in most cases, the cause of the disease is not fully understood. Interestingly, standard of care has recently evolved from full thickness corneal transplants to procedures that preserve as much of the host cornea as possible with spectacular outcomes. As an unexpected result of donor detachment complications of these surgeries in some cases, the progressive recovery of the corneal transparency suggests a possible re-endothelialization of the central cornea by the host cornea. Moreover, there are now case studies that report that simply removing the central host corneal endothelial membrane, without donor tissue transplantation, has lead to corneal clearing, evidence that there are host stem/progenitor cells capable of proliferating and migrating toward the central cornea during the healing process. The goal of this research program is to investigate the potential cell regeneration in the corneal endothelium and the mechanism by which this process may be compromised in case of FECD. 

The first part of this presentation will highlight the capabilities of a cellular resolution Gabor-domain optical coherence microscope (GD-OCM) that was developed to investigate corneal disease and to advance our understanding of FECD.

The second part of the talk will discuss the regenerative capabilities of corneal endothelial cells as an alternative therapy for the future treatment of endothelial dystrophy.

9/16/2016 Hannah Block Reciprocal effects of change in visuo-proprioceptive and motor processing
The brain can estimate hand position both visually, from an image on the retina, and proprioceptively, from sensors in the muscles and joints. The brain is thought to make best use of available sensory estimates through multisensory integration, giving us the flexibility to cope with the frequent sensory perturbations we experience: E.g., realigning one or both sensory estimates when they become spatially misaligned, as when washing dishes with the hands immersed in water, which refracts light. Importantly, any change in the brain’s perceptual estimate of hand position will naturally influence all movements planned with that hand. Despite this fundamental link, these processes have traditionally been modeled independently. We are investigating whether perturbing components specific to multisensory models affects components specific to motor control models, and vice versa. For example, does visuo-proprioceptive realignment alter excitability in primary motor cortex, a substrate of motor control with no known multisensory role? Results from human behavior and neurophysiology suggest such an effect, which has important implications for our understanding of both multisensory integration and motor control.
9/23/2016 Xiao Fu Progression of Diabetic Capillary Occlusions: An Explanatory Computational Model Early stages of diabetic retinopathy often feature contiguous areas of capillary occlusions in both perifoveal and peripheral areas. Vascular endothelial growth factor (VEGF)  is thought to be physiologically integral to the leukocyte-endothelial adhesion that initiates  individual capillary closure. However, the underlying mechanisms by which large ischemic areas develop from single capillary  occlusion are not clear. To explain the observed pattern of contiguous capillary blockage seen in diabetics, we developed an explanatory computational  model based on a hypoxia-driven vascular endothelial growth factor (VEGF)-mediated  adverse feedback  mechanism. Our main hypothesis is that hypoxic retinal tissue near an occluded capillary produces VEGF, which diffuses to adjacent patent capillaries and increases  their probability of occlusion.  Methods: We computationally constructed a network of retinal capillaries in the perifovea (with an architecture derived from a human perifoveal arteriovenous sector capillary map obtained using AOSLO imaging) and in the retinal periphery. We calculated blood flow rates in the constructed capillary network using a hemodynamic model. We simulated progression of capillary occlusion, distribution of oxygen and VEGF, and formation of edema.  Results: Our simulations in the perifovea produced a pattern of spreading capillary loss with  associated macular edema. Our simulations in the retinal periphery showed spreading ischemia resulting from the progressive loss of the ladder capillaries which connect peripheral arterioles and venules. In perifoveal simulations, the temporal evolution of blood flow agreed quantitatively with experiment, and the venous oxygenation agreed qualitatively. Replicate simulations in the perifovea showed that different sites of the initial capillary occlusion would result in different sized areas of retinal ischemia.  Conclusions: We built an explanatory computational model that plausibly reproduced clinically manifested contiguous loss of capillaries in perifovea and retinal periphery at early stages of diabetic retinopathy. Our simulations implied that different sites within the retinal capillary network could have different propensities for progression of capillary occlusion and ischemia. Future computer simulations may allow precise, individual-based laser photocoagulation applied to those areas of retina most likely to become ischemic as an intervention early in the course of diabetic retinopathy, and help reduce the risk of blindness by preventing propagation of capillary occlusions.
9/30/2016 Rob deRuyter

Using bilocal visual noise to investigate motion detection in flies and humans.

We study wide field motion detection using an apparent motion stimulus that has motion energy only at proscribed locations in space and time. This allows us to disentangle some aspects of the computational structure of motion that are hard to separate with more conventional stimuli. I will first present results from a set of experiments based on electrophysiological recordings from fly motion sensitive neurons (Roy and de Ruyter van Steveninck, Journal of Vision 16:8, 1-19 (2016)). That will be followed by a discussion of some preliminary data based on human psychophysics, which indicate that some of the same adaptations may be at work in both species.

10/14/2016 Artur Cideciyan

Gene Therapy for the Improvement of Vision and Arrest of Retinal Degeneration

Gene therapy is hailed as a potential treatment for many inherited retinal degenerations caused by single gene defects that cause severe blindness. In human clinical trials to date there has been convincing evidence that gene therapy can quickly improve aspects of vision over a matter of days to weeks. More controversial is whether gene therapy can also arrest slow degeneration of photoreceptors which typically occurs over years to decades. Two examples will be presented and results of gene therapy discussed..
10/28/2016 Alfred Lewin

Gene Delivery of Secreted Cell Penetrating Peptides to Prevent Retinal Degeneration in a Mouse Model of Oxidative Stress in the Retinal Pigment Epithelium

Damage caused by reactive oxygen species is believed to contribute to the damage to the retinal pigment epithelium (RPE) and choroid leading to the development of age related macular degeneration (AMD). To model this process in a short-lived animal and to test the hypothesis that mitochondria are a source of RPE oxidative stress, we generated a mouse model in which Sod2, the gene for mitochondrial superoxide dismutase, is deleted in the RPE. These mice exhibit some of the salient features of dry AMD including accumulation of lipofuscin, disorganization of Bruch’s membrane, atrophy of RPE cells and death of associated photoreceptors. We have used this model to test gene therapies for geographic atrophy, the advanced form of dry AMD.  Adeno-associated virus (AAV) vectors were designed to deliver secreted, cell penetrating peptides to block the activation of the NLRP3 inflammasome or to induce the production of antioxidant enzymes in the RPE.  These vectors reduced inflammatory stress in the retina in the endotoxin induced uveitis model and protected the retina in an acute model of oxidative injury caused by injection of sodium iodate.  AAV expressing a caspase activation and recruitment domain (CARD) peptide reduced the rate of retinal degeneration in the Sod2 deletion model of geographic atrophy.  We anticipate that these gene delivery approaches will have benefit in preventing advanced dry AMD in human patients.
11/4/2016 Student presentations

Pre- American Academy of Optometry talks

This is for the students to present the work they will be taking to the AAO meeting next week.  The hope is that this session will provide a good practice with good individual feedback for the presenters, and provide us all opportunity to see what will be going on at AAO (and these presentations) if we cannot otherwise attend.  Paper talks will mirror the format at AAO, and posters will be presented electronically on the screen.


SPEAKER:   Student presentations  (*=paper)


Teddy Chow*


Edmund Arthur*


Debra Antwi


Basal Altoaimi


Alice Grasso


Casey Car


program end

11/18/2016 Mallika Valapala

Induction of Lysosomal Function for the Treatment of Age-related Macular Degeneration

Age-related macular degeneration (AMD) is a progressive degenerative disease of the eye and is the most common cause of visual loss among the elderly. Atrophic degeneration of retinal pigment epithelial (RPE) cells is a major contributor to the pathogenesis of AMD. Recent evidence suggests that in AMD, reduced lysosomal activity is associated with accumulation of lipofuscin and intracellular lipids resulting in progressive degeneration of the RPE. Lysosomes are crucial for proper functioning of the RPE, since these cells are subjected to the daily burden of phagocytosis of photoreceptor outer segments (POS) and autophagic degradation of cellular metabolic waste. The mechanisms that regulate lysosomal function and thereby influence autophagy in the RPE are not well understood. The focus of my laboratory is to identify molecular mechanisms regulating lysosomal activity and to use this knowledge to develop effective therapeutic strategies for the treatment of AMD.  Our studies suggest the importance of ZKSCAN3 (zinc-finger protein with KRAB and SCAN domains 3) and transcription factor EB (TFEB) as master regulators of lysosomal function and autophagy in the RPE. We provide evidence to show that ZKSCAN3 and TFEB-mediated transcriptional program induces flux through lysosomal degradative pathway in the RPE. Our studies are directed to exploring the in vivo potential of ZKSCAN3 and TFEB in alleviating abnormal substrate accumulation and degenerative loss of RPE.

12/2/2016 Paul Gamlin Intrinsically-photosensitive Ganglion Cells: Anatomy, Physiology, and Behavioral Roles


Primate vision starts with the activation of rod photoreceptors in dim light and short (S)-, medium (M)-, and long (L)- wavelength sensitive cone photoreceptors in daylight. A parallel, non-rod, non-cone photoreceptive pathway, arising from a unique population of melanopsin-expressing, intrinsically-photosensitive retinal ganglion cells (ipRGCs), has been demonstrated in mammals including primates. Our understanding of the anatomy, physiology, and behavioral roles of primate ipRGCs has improved substantially over the past decade. In primates, this anatomically distinct population of ipRGCs is strongly activated by rods and cones, and the intrinsic light response combines with rod and (L + M) cone-derived responses to signal irradiance over the full dynamic range of primate vision. Each of these component ipRGC light responses has distinct spectral and temporal attributes that can be characterized both in vitro and in vivo. Further, in primates including humans, we have demonstrated that these cells play an important role both in light-evoked pupillary responses as well as in the sustained pupil constriction that occurs after light cessation – the post-illumination pupil response. The post-illumination pupil response is driven by the intrinsic photoresponse of ipRGCs, and can be used to evaluate ipRGC activity in normal and diseased retinas.  

12/09/2015 Stephen Burns



Ann Elsner

Individual Differences in Human Retinas: Neuroplasticity in Aging and Age-related Macular Degeneration and the Potential for Cone Survival and Vision


Large individual differences in cone densities occur even in healthy, young adults with low refractive error. We developed a model of how cone density changes over the retina for young individuals, as compared with older individuals and patients with age-related macular degeneration.  Some individuals have more cones than others, but this change is not uniform over the retina. The individual differences are related to gender and photoperiod, as well as aging and retinal disease.  The coefficient of variation was 0.0767 for total cones, but higher for samples near the fovea. Individual differences occur both in total cones and other developmental factors related to cone distribution. The cones in older eyes are not only fewer, but also shorter.  Cones in some regions are relatively spared, possibly because they have relatively more metabolic resources in their environments.  This finding, taken together with the findings of cones that live without normal retinal pigment epithelium, indicate that neurons alter their structure that allows a more economic existence, which may help explain our surprising finding of the survival of sick cones for years in patients with age-related macular degeneration.  Cones, therefore, provide a target for therapies to preserve vision.


Debra Antwi

Brad Caron

Measurement of fluorescein concentration in the Tear Film by image analysis and tear sampling

The effects of concussion-prone athletics on the brain:  A comparison of football, cross-country, and socioeconomically matched non-athletes using fMRI and diffusion-imaging.

Tear film hyperosmolarity and tear break-up (TBU) are thought to be factors promoting ocular surface stress in the development of the dry eye condition.  Theoretically, as the tear film evaporates and TBU occurs in the interblink interval, tear film osmolarity should increase locally and/or globally. Mathematical estimates and indirect measures of this increase suggest levels as high as 1500mOsm/Kg (sea water is ~1100mOsm/Kg) within areas of TBU.  However, there is currently no method for measuring dynamic changes in tear film osmolarity over the ocular surface.

We propose to use changes in tear film fluorescence during the interblink interval as a surrogate for modeling dynamic changes in tear film osmolarity.  As a first step in this process, we have developed a method for determining the concentration of fluorescein in the inferior meniscus by the intensity of fluorescence.  The method has been calibrated with known concentrations of fluorescein and tear samples were taken and analyzed to check the measurements.  Once the concentration of fluorescein is known in the inferior meniscus, we will apply it to the initial fluorescence of the tear film over cornea in the interblink interval.  Knowledge of FL concentration within the tear film before TBU begins allows better estimation of relative changes in tear film thickness and estimation of osmolarity during TBU.  We will present this calibration data and other preliminary findings.

Football players receive a high number of impacts throughout a season that may or may not produce a diagnosed concussion.  McAllister et al (2014) found that Division I football players without a concussion have over 500 impacts of 54Gs or higher, on average, for a season.  Research into the effects of these sub-concussive impacts has shown neurocognitive, functional MRI, and white matter differences.  I will be presenting research performed here at IU investigating the effects of sub-concussive blows on the brains of varsity Division I football players.  4th and 5th year varsity IU football players, age-matched cross-country runners, and socioeconomically matched non-athletes underwent a functional MRI smooth pursuit task (as smooth pursuit has been shown to be impaired in concussed patients (Mucha et al, 2016)) and diffusion MRI.  Functional and anatomical differences between the brains of the football players and the other two groups were observed.


Senyo Bright

Amos Wang

Mitigating the impact and understanding the implications of the presence of putative proliferated glia in glaucomatous OCT en face imaging

Measuring the axial power of multifocal contact lenses using Shack-Hartmann aberrometer

The disagreement between structural and functional measures in the diagnosis and management of glaucoma (referred to as structural-functional discordance) poses a diagnostic dilemma to clinicians. Amongst other things, the spatial challenges in relating structural measures to functional measures have been identified as a source of the structural-functional discordance. Also, the poor spatial sampling of the 24-2 perimetric protocol of the macula and the entire retina have also been identified as a source of the structural-functional discordance.

En face imaging technology enhances visualization of the retinal nerve fiber bundles (RNFBs), and their spatial locations relative to other retinal anatomical landmarks. Given the RNFB detail revealed by OCT en face imaging, it holds promises to overcome the spatial concerns in relating structural measures to functional measures. It also holds promises for improving the sampling of the perimetric testing through targeted perimetry.

The levels of RNFB detail revealed by OCT en face imaging is however sometimes obscured by putative proliferated glia, referred to elsewhere as activated retinal astrocytes and Muller cells (ARAM). The reflecting ARAM poses challenges to visualizing or quantifying the RNFB in enface imaging. Also, ARAM has been suggested to be due to neural remodeling in response to neural insult.

We investigated the potential of using depth parameters of RNFL en face imaging to mitigate the impact of reflecting ARAM on visualizing or quantifying the the RNFBs. We also investigated the implication of the presence of ARAM on OCT en face images to glaucomatous neural degeneration.

Multi-zone contact lenses have been used to expand depth of focus for presbyopic eyes and to control eye growth in myopic eyes. Because of abrupt multiple power changes associated with many multi-zone contact lenses, a sampling-based Shack-Hartmann (SH) aberrometry (e.g., 150mic for ClearWave) may possess significant challenges. To discover these challenges, I  measured the power profiles for the following lenses: (1) Two “true” zonal bifocals with known power profiles (+2D trial lens with a central 3mm hole; and -2D trial lens with a central 3mm hole);  (2) A two-zone commercial multifocal contact lens (AirOptix with high Add); (3) A center near three-zone commercial multifocal contact lens (Biofinity high Add); (4) A center distance three-zone commercial multifocal contact lens (Biofinity high Add); (5) A four-zone multifocal contact lens (Oasys with high Add); and (6) A myopia control lens (three-zone Misight with high Add). I used different methods to compute the axial power of these lenses: (1) 4th order polynomial fit; (2) 10th order polynomial fit; (3) local slope method; and (4) grouping of individual zone (multi-zone method). I will discuss the issues of each method, and compare my results with previously reported multifocal powers using different non- Shack-Hartmann techniques, and provide the best candidate measure.


David Crandall, PhD

Egocentric computer vision, for fun and science

 Low-cost wearable cameras are now entering the mainstream, allowing people to record their lives from first-person or "egocentric" perspectives that approximate people's own first-person fields of view. In addition to letting people create visual "life-logs", they may enable more serious applications, like providing memory aids for dementia patients, serving as assistants for the visually-impaired, or enhancing accountability when worn by police officers. However, these cameras also raise challenges, including the huge volume of imagery they produce (typically thousands of photos or dozens of hours of video per day), and the significant privacy concerns they create. I'll describe two lines of collaborative work on using computer vision to automatically understand and manage first-person imagery. The first is for consumer applications, where our goal is to develop automated classifiers to help categorize lifelogging images across several dimensions. The second is using computer vision with wearable cameras to people's interactions with others and the environment to better understand how they learn. Despite the different goals, these applications share common themes of robustly recognizing image content in noisy, highly-dynamic first-person imagery, and provide a motivating backdrop that I'll use to review the dramatic progress in computer vision over just the last few years.

For more information, please visit:   http://vision.soic.indiana.edu/


Kathy Sapoznik

Ting Luo

Multiply-scattered light AOSLO imaging using a configurable aperture approach

Modeling retinal arterial diameters and velocities at bifurcations

Multiply scattered, non-confocal imaging has been utilized in retinal imaging systems for many years. In adaptive optics retinal imaging systems, this approach has led to the visualization of retinal vessel structure and cone inner segments.  In our new adaptive optics scanning laser ophthalmoscope (AOSLO) we have the ability to change the shape of the aperture numerous times within one imaging session utilizing a micro-mirror array (TI-DLP 6500). MATLAB is used to upload various shapes onto the array and the on state of the array is sent to one detector while the off state is sent to another. At this stage we have compared imaging with a split detection arrangement orienting the split at 0 and 45 degrees and a split annulus rotating in 45 degree increments with different radii. Varying the rotation of the split annuli enables enhanced visualization of retinal vessel walls and oriented flow orthogonal to the annulus orientation. A configurable aperture has the ability to change from frame to frame and should allow imaging the retinal cellular features in a controlled manner.

Murray’s law describes the vascular requirement of tissues based on minimizing energy requirements for blood distribution. In this study we examine the predictions of Murray’s law by including the variation in viscosity with vessel size (the Fåhræus-Lindqvist effect) and derived the interaction of viscosity, vessel size, and blood velocity at vascular branches. Size data were measured in 26 healthy subjects (203 arteriolar bifurcations) using confocal and multiply scattered light AOSLO images. Velocity data were from 3 normal subject (30 arteriolar bifurcations) using AOSLO temporal offset imaging. Size and velocity data were averaged and fit using the Matlab curve fitting toolbox.  This modeling predicts that the exponent in Murray’s law varies from 2 to 3 depending on vessel size. Vessels smaller than 40μm would have an average of 2.24, larger vessels would be close to the classical prediction of an exponent of 3. The diameter measurements gave an exponent of 2.15 (1.89-2.41, 95% confidence interval) for vessels less than 40μm and an exponent of 2.74 (2.54-2.94) for vessels between 40 and 100μm. Velocity data for small vessels (<30μm) had an exponent of 2.56 (2.08-3.04). This model may allow better understanding of the relation between the changes in vascular branching with diseases such as diabetes which effect the local viscosity of blood (which includes both hematological and other local factors, such as vascular walls) in the living human retina. 


Hin Cheung

Yen-Chiao Wang

Potential biomarker for glaucoma-related damage to the retinal nerve fiber layer

Spdef regulates the early stage development of conjunctival goblet cell

Glaucoma is the most common optic neuropathy, characterized by progressive loss of retinal ganglion cells, cupping of the optic nerve, and visual field defects. Historically, the clinical focus has been on examination of the optic nerve head with stereo biomicroscopy and visual fields, but newer technologies such as spectral domain optical coherence tomography (SD-OCT) and adaptive optics scanning laser ophthalmoscopy (AOSLO) are providing new insights into this disease. Using the AOSLO, glaucoma patients with clear wedge defects and controls were imaged around the optic nerve head at the level of the retinal nerve fiber layer (RNFL). Small, hyper-reflective bodies can be seen within the axonal bundles in healthy RNFL. The number and reflectivity of these bodies significantly decrease in areas with RNFL damage. One theory is that these small, hyper-reflective bodies may be mitochondria, and as axons become damaged, the function of the mitochondria is compromised and these organelles also begin to die. This may be a potential biomarker to identify retinas at risk for glaucoma-related vision loss. 

Purpose: To study the effect of spdef in early stage development of conjunctival goblet cell. Previous study suggested that 8 months old Spdef knock out mice lack goblet cell in conjunctiva. However, the possible mechanisms of such phenotype remain unknown. In this study, animal experiments were performed to investigate the possible mechanisms.  Methods: Spdef knock out mice were used to investigate the whole process of goblet cell development. Compare to the lack-of-function, we will use tet-o-spdef transgenic mice to perform the gain-of-function to study the critical timing of rescue. During the whole experiment, HE staining, PAS, conjunctival whole mount and MUC5AC immunostaining will be performed.   Results: In third week, almost all goblet cells cannot develop in spdef knock out mice. However, we still can see very small amount of goblet cell analogue.  Conclusions: Spdef plays a key role in conjunctival goblet cell development. However, the result indicates a possible thinking, there is a new signal pathway other than spdef signal pathway, can control goblet cell development.


Furu Zhang

Arthur Edmond

Diurnal Time Course of Cone Photoreceptor Disc Shedding in the Living Human Eye

Neural and vascular changes in the eyes of diabetic patients

Purpose: Photoreceptor outer segments (OS) undergo periodic shedding of small packets of membranous discs at their distal ends. This process, called disc shedding, is fundamental to maintaining health of the outer retina and believed to follow a diurnal rhythm. While disc shedding has been studied extensively in animal models, little is known in humans. In this study we took advantage of our recent discovery of the optical signature of disc shedding in the living human eye using adaptive optics-optical coherence tomography (AO-OCT) [1] to investigate the diurnal properties of this elusive event.  Method: Two healthy subjects free of ocular disease were imaged at 3° temporal retina (1.5°×1.5° FOV) with the Indiana AO-OCT system. Subjects were entrained to a 16/8 hour awake/sleep cycle for three days preceding the experiment. On the day of the experiment, the subjects awoke 2 hours before the entrained wakeup time, and the experiment started 1 hour later. Volume videos were acquired every 15 minutes for 17 hours. One subject stayed under stable room light during the entire experiment, while the other was also exposed to three 10-minute intervals of natural light over the day. B-scan and average A-scan time traces of each tracked cone were generated with the post-processing method described by Kocaoglu et.al [1]. Both time-trace presentations were used to detect disc shedding events as manifested by spatial (axial) and temporal changes in inner segment/outer segment junction and cone outer segment tip reflections. In one subject, the OS length of each cone at each time point was measured.   Results: Thousands of cones were successfully imaged and tracked over the 17 hour period in both subjects. Shedding events were detected in 79.5% and 72.6% of the tracked cones. Similar to previous animal studies, shedding prevalence exhibited a diurnal rhythm. But unlike almost all other diurnal animal models studied, shedding peaked shortly after the entrained wakeup time. Consistent with this, traces of the average cone OS length across the 17 hours implied maximum shedding shortly after the entrained wakeup time. Distinct from these diurnal changes, a significant but transient increase in shedding followed exposure to natural light.  Conclusion: Human cone photoreceptor shedding exhibits a diurnal rhythm that peaks in the morning and is sensitive to natural light.

Tissue oxygenation of the inner retinal layers is provided by the inner retinal capillaries, and the relatively constant inner retinal layer thickness at the margins of the foveal avascular zone has been attributed to the limits of the metabolic support by these capillaries (Chui et al, IOVS 2014). Disruption of parafoveal capillary network and hence a large foveal avascular zone size has been found in type 2 diabetic patients prior to the clinical detection of diabetic retinopathy using Adaptive Optics Scanning Laser Ophthalmoscope and Optical Coherence Tomography Angiography (Tam et al, IOVS 2011, Choi et al, Retina 2016). We investigated whether there is a change in the inner retinal layer thickness at the margins of the foveal avascular zone in response to the changes in the parafoveal capillaries and the foveal avascular zone size in diabetic patients who clinically have not been diagnosed as having diabetic retinopathy. To quantify these potential changes, we used en face Optical Coherence Tomography to obtain foveal avascular zone area, foveal avascular zone effective diameter, and inner retinal layer thickness at the margins of the foveal avascular zone of diabetic patients without diabetic retinopathy or only mild non-proliferative diabetic retinopathy and their age- and gender – matched controls. Our results suggested that even though early in the course of retinal changes with diabetes, there is retinal vascular remodeling at the foveal avascular zone: the inner retinal layer had not thinned to decrease the neural mass at the margins of the larger foveal avascular zone. Further, we have previously shown that diabetic males with clinically significant macula edema have a greater than 12 micron thicker central macular thickness as compared to diabetic females with clinically significant macula edema (Young et al, AAO 2014), consistent with conclusions that compared to diabetic males, younger diabetic females have a form of neuroprotection from estrogen (Ozawa et al, IOVS 2012). We investigated which individual retinal layers in these diabetic males are thicker as compared to the diabetic females. To address this, we quantified the individual retinal layer thicknesses of these diabetic patients to show the effect of gender, eccentricity, and the interaction between these two independent variables. We found that outside the fovea in diabetic patients with clinically significant macula edema, the nerve fiber layer and ganglion cell layer-inner plexiform layer thicknesses were significantly greater in males than females.

3/3/2017 Martin S. Banks TBA





Ayoub Lassoued

Essam Almutleb

Classification of cones across the transition zone in retinitis pigmentosa based on AO-OCT signature

Street-Crossing Decision-Making in Pedestrians with Simulated Central Field Loss

Purpose :  Retinitis pigmentosa (RP) is the most common form of inherited irreversible visual loss worldwide.  To better describe photoreceptor loss in human retina, we developed a cone classification scheme based on the longitudinal reflectance profile of cones imaged with adaptive optics optical coherence tomography (AO-OCT).  Method :  We used the Indiana AO-OCT system at 500 KHz A-scan rate to image two subjects: one with late onset adRP and one control between 1.5 and 7.5 degree. Based on the cross-sectional profile of individual cones, cone outer segments (OSs) were classified as containing two or more reflections (interpreted as normal), containing one reflection (interpreted as degenerating at one tip), or containing no reflection (interpreted as degenerating at both tips). To capture the variability in OS morphology, standard deviations in 2-reflection cones’ OS length and axial center position were computed.  Results :  After examination of averaged A-scans over individual cones, normal OSs were found to constitute more than 89% of the total in the normal subject  at any eccentricity, while in the RP subject, the percentage decreased with eccentricity from 77% at 1.5° to 0% at 7.5⁰. 2- reflection cones for the RP subject exhibited both a higher variation in the OS length (SD=4.5-8.7 μm ) and center position (SD=0.5-3.3 μm ) than for the normal subject (SD=2.3-4 μm  and SD=0.1-0.4 μm respectively). The percentage of 1-reflection OSs was less than 9% at all locations in the normal retina; in the RP retina, it increased from 22% at 1.5° to 48% at 4.5° then decreased gradually thereafter. The percentage of 0-reflection OSs was less than 2% in the normal retina, while it increased from 0% at 1.5⁰ to 63% at 7.5⁰ in the RP subject.  Conclusion :  AO-OCT reveals RP cone OSs to have reduced number of reflections and more varied length and axial location.

Street-crossing decision-making is negatively affected by vision loss. This study tested the hypothesis that as the size of an experimentally induced central scotoma increases, street-crossing performance worsens.  Methods: Street-crossing decisions were measured in twelve young subjects aged between 23 and 31 years while monocularly viewing a non-signalized one way street for different vehicular arrival times (0.3-17 sec). Using a 5-point rating scale, subjects judged whether they could cross the street prior to vehicular arrival with habitual vision and simulated central field loss (CFL) using contact lenses (CLs) with different central opaque diameters. Using Receiver Operating Characteristics (ROC) curve analysis, we obtained subjects’ accuracy (amount of time in seconds where subjects either over- or under-estimated vehicular arrival time relative to their actual crossing time) and reliability (how quickly subjects transitioned from judging insufficient to adequate time to cross relative to their actual crossing time). Underestimation of vehicular arrival time relative to subjects’ crossing time generated positive inaccuracies but safe crossing judgments the converse was true for negative values. The quicker the transition from judging insufficient to adequate time to cross, the more reliable and hence better performance. The converse was true for slow transitions.  Results: The centrally opaque CLs induced central scotomas with mean diameter of 18.85 degrees (±7.37 degrees). The mean accuracy values in street-crossing decisions for normally-sighted people with and without a simulated central scotoma were 1.43±0.27 sec and 1.40±0.47 sec, respectively. No significant difference in accuracy was found between both conditions (p= 0.780). The median of reliability in street-crossing decisions was 0.55 for both conditions. Surprisingly, no statistically significant correlation was detected between scotoma size and accuracy and reliability in subjects’ street-crossing decisions. (p=0.567 and p=0.265, respectively).  Conclusion: This data suggests that normally-sighted people with and without a simulated CFL adopted a safe street-crossing strategy by underestimating the vehicular arrival times relative to their crossing time. Moreover, an induced central scotoma did not appear to negatively affect the accuracy and reliability of subjects’ street-crossing decisions.



Sonisha Neupane

Ray Warner

The relationship between motor behavior and perception in strabismic subject

Flicker evoked changes in retinal capillary blood flow measured using a dual channel Adaptive Optics Scanning Laser Ophthalmoscope


Strabismus is prevalent in 2- 4%. of the general population.  When constantly present, strabismus disrupts the binocular visual system. It may be accompanied by a combination of amblyopia, anomalous retinal correspondence, diplopia, eccentric fixation and/or suppression. But when the deviation is intermittently present, there is the potential for good visual performance during periods alignment, but the need to manage diplopia during misalignment. The visual perception of intermittently strabismic subjects in the real world when their eyes are aligned or deviated is almost consistent, but their motor and sensory status are vastly different. This talk will examine the relationship between the sensory, motor and perceptual status in strabismic subjects.

Neurovascular coupling has been measured in the human eye using a variety of techniques. However, until recently it was not possible to directly measure changes in blood flow at the capillary level. We use a dual channel Adaptive Scanning Laser Ophthalmoscope (AOSLO) to measure capillary blood velocity in the human retina with visual stimulation. The retina of four healthy subjects from ages 23-62 were imaged with the Indiana dual channel AOSLO. Channels were offset 50 lines vertically to produce a temporal separation of 3.2 ms. Capillaries around the fovea were imaged simultaneously in both channels, each using a 500μm offset aperture, in three sessions: (1) No flicker (Uniform), (2) Full-Field Flicker at a 20 Hz rate and (3) Uniform again, resulting in approximately 8 hundredframe- videos (at 33Hz) for each subject under each condition. Red blood cells (RBC) traveling through selected capillaries were detected using a local-temporal variance mask, and the local average were computed within a capillary segment over 3 frames. The same frames in the second channel were averaged to measure the displacement and thus the velocity (mm/s). The velocity of RBCs were computed over time and compared between sessions to measure the effect of onset/offset flicker. Capillary blood velocity measurements were possible in all four subjects. Individual capillaries varied in velocity and pulsatility. The speed of RBCs under Full-Field Flicker condition increased in all subjects, on average from 2.3mm/s under uniform field stimulation (session1) to 2.7mm/s during flicker stimulation (session2). Following offset flicker stimulation (session3), the velocity decreased back to 2.2 mm/s. Average velocity under the flicker stimulation is significantly higher than those under the uniform field after visual stimulation (p <0.05; one-tailed t-test with 95% confidence interval). The use of a dual channel AOSLO allows measurements of RBC velocities in retinal capillaries. Visual stimulation produces measurable differences in velocity, although individual capillaries differ in their response within the same region. The temporal resolution can provide insight into the pulsatility and average blood flow in retina capillaries.


Note:  Oxyopia will end by 12:30 as we only have 1 student speaker.

Moonjung Choi

Functional roles of SLCA411 and mitochondria will influence our understanding of how the loss of SCL4A11 results in corneal endothelial dystrophies

Corneal endothelial cells contain a large number of mitochondria to generate the energy that is needed for the active transport of water and ions from the stroma to maintain corneal deturgescence. Mutations in the SLC4A11 gene are associated with congenital hereditary endothelial dystrophy type 2 (CHED2) and late-onset Fuchs endothelial corneal dystrophy (FECD). The human SLC4A11 gene encodes three N-terminal transcript variants. SLC4A11-B and SLC4A11-C are localized to the plasma membrane, but SLC4A11-A is localized intracellularly. Our lab has shown that loss of SLC4A11 produces mitochondrial dysfunction. Therefore, we ask if SLC4A11-A is associated with mitochondria. Mitochondria were isolated from SLC4A11+/+ and SLC4A11-/- mouse corneal endothelial cell lines. SDS-PAGE protein staining and Western Blotting will be performed to show if SLC4A11 is in the mitochondria and compared to the expression of SLC4A11 in whole cell or plasma membrane lysates. Determining the functional roles of SLCA411 and mitochondria will influence our understanding of how the loss of SCL4A11 results in corneal endothelial dystrophies.

4/7/2017 Nancy McNamara




Colin Downey

Yifei Wu

Dynamic Characteristics of 5 to 22 week-old Infants’ Accommodation and Vergence Tracking Responses


The motor and sensory binocular visual systems develop dramatically during the first 5 months after birth.  Typical development of stereopsis is dependent on adequate image quality and alignment in the dynamic natural environment, requiring appropriate accommodation (A) and vergence (V).  This study documented the development of dynamic characteristics of infant accommodation and vergence responses between 5 and 22 weeks of age, using a correlogram approach (Mulligan et al, 2013).  Fifteen infants (two excluded due to restlessness) and 10 adult controls were tested during trials in which they viewed naturalistic cartoons presented on a screen. The stimulus moved as a random walk in depth between 33-80cm for 100 seconds. Horizontal eye position and refractive state were recorded binocularly and simultaneously at 50Hz using eccentric photorefraction and Purkinje-image eye-tracking analyses of video images (PowerRef 3, PlusOptix). Cross-correlations of the stimulus and response velocities demonstrated steady maturation of (V) and (A) between 5 and 22 weeks. Adults demonstrated mean peak correlations of 0.80±0.03 for (V) and 0.75±0.10for (A, left eye), at mean time lags of 0.15s±0.04 (V) and 0.20±0.05 (A,left). 5-9 week-olds, 10-13 week-olds and >14 week-olds had mean peak correlations of 0.54±0.17, 0.67±0.11 and 0.60 ± 0.15 (V) and 0.52±0.16, 0.61±0.12, 0.52±0.14 (A,left), at time lags of 0.42s±0.20, 0.25s±0.07, 0.26s±0.11 (V) and 0.36s±0.12, 0.23s±0.07, 0.28s±0.13 (A, left).  These data indicate that by five weeks of age, some infants are able to track these unpredictable dynamic stimuli with short delays on the order of only 0.3s (V) and 0.2s (A, left) slower than adults, albeit with more response variability as indicated by reduced peak correlations.



Alice Grasso

Maddie Jones

A Comparison of Stereotests

The Mind Can't Go Blind: How Age-related Macular Degeneration Affects Street Crossing Time Estimates

Stereo vision is an important aspect of normal binocular function and should be assessed for each patient. Stereotests are used in a variety of clinical ways, from screening for binocular vision abnormalities to gauging efficacy of treatments for strabismus and amblyopia. Effective stereotests are therefore a necessary part of any optometrist’s practice.  Current clinical stereotests have limitations, especially for younger children. Current test books are not engaging enough to keep young kids interested for an extended period of time. Current tests require a fixed testing distance to be accurate, and they place patients’ stereo thresholds into categories, rather than determining the exact stereo threshold. Often they are solvable monocularly, though stereotest stimuli are supposed to be solvable binocularly only. New stereotests are attempting to correct these limitations and so be more usable for eye health screening and for use in pediatric optometric practice.  I will be presenting my MS Thesis project, which compares four stereotests in terms of accuracy and test-retest variability.  I will present and compare the clinical standard Randot stereotest, the Randot Preschool stereotest, a newly developed stereotest for use on a glasses-free 3D tablet, and another stereotest (like that for the 3D tablet) that is done using a 3D projector.

Purpose: The purpose of this study was to determine if Age-related Macular Degeneration (AMD) impacted a person’s ability to estimate their own crossing time relative to their actual crossing time compared to age-matched subjects with normal vision.  Methods: 32 AMD subjects and 32 age-matched, normally-sighted subjects (controls) estimated their time to cross a single-lane, one-way street four times (pre-estimate). Then subjects actually crossed the street four times, and subsequently estimated their crossing time four additional times (post-estimate). A linear mixed model with repeated measures for subject was used to determine if the ratio between subjects’ estimated and actual crossing times changed as a function of subject group (AMD vs. control) and whether estimates changed before and after actually crossing the street.  Results: No significant difference was found between AMD and controls in their crossing ratios (p = 0.54). This was true for both the pre- and post-crossing ratios (p = 0.96).  Both subject groups’ crossing ratios however decreased significantly going from pre to post (p < 0.0001). Specifically, before and after crossing the street, the reduction in crossing ratios were on average 12.9% and 13.2% for the AMD and controls respectively.  Conclusion: Our data suggests that familiarity with the street as opposed to central vision loss from AMD affects a person’s ability to estimate their crossing time.


Kazuhiro Kurokawa