Faculty Bibliography

Purpose : To evaluate the extent of anatomical changes of the retinal ganglion cell axons and trans-neuronal changes in the optic radiation across glaucoma stages using noninvasive MRI and clinical ophthalmic assessments. Methods : This observational, cross-sectional study included 21 advanced glaucoma, 16 early glaucoma, and 13 healthy subjects who underwent spectral-domain OCT imaging of the eye, 3-Tesla anatomical MRI, diffusion tensor imaging (DTI) of the brain, and perimetry. Glaucoma staging criteria were based on recommendation by the American Glaucoma Society ICD-10 coding guidelines. Optic nerve and optic chiasm volumes were manually estimated from anatomical MR images. DTI-derived parametric values [fractional anisotropy (FA), mean diffusivity (DTI-MD), axial diffusivity (AD), and radial diffusivity (RD)] were extracted using manual regions-of-interests drawn on the optic radiation of each hemisphere. The anatomical MRI and DTI parameters were compared to clinical OCT parameters [peripapillary retinal-nerve-fiber-layer (pRNFL) thickness, macular ganglioncell- inner-plexiform layer (GCIPL) thickness, optic nerve head cup-to-disc ratio (C/D)], visual field mean deviation (VF-MD), and to each other using linear mixed-effects models. Each parameter was also compared across groups using one-way MANOVA and receiver operating characteristic (ROC) analyses. Results : From linear mixed effects models, optic nerve and optic chiasm volumes were associated positively with pRNFL thickness and VF-MD (p<0.05), but not GCIPL thickness or C/D. Optic chiasm volume was significantly associated with FA, DTI-MD, and RD, while optic nerve volume was negatively associated only with RD (p<0.05). Summary statistics in Figure 1 indicated significantly thinner pRNFL and higher C/D in early glaucoma than healthy control, whereas volumetric brain measurements, optic radiation DTI parameters (FA and RD), and VF-MD altered significantly between early and advanced glaucoma. ROC analyses in Figure 2 indicated differential abilities among MRI and clinical ophthalmic techniques to distinguish between glaucoma stages. Conclusions : Both volumetric and diffusional brain changes measured from anatomical MRI and DTI may be useful for examining glaucomatous damages along the visual pathway complementary to OCT and perimetry

Purpose: Currently, lowering intraocular pressure (IOP) is the only clinically proven way to treat glaucoma. However, glaucomatous neurodegeneration may still continue after IOP is reduced. Recent studies suggest that citicoline may improve visual outcomes in glaucoma patients, yet its underlying mechanisms remain poorly understood. Here we examined the effects of IOP elevation on visuomotor response and brain functional connectivity (FC) with and without citicoline treatment.
Method(s): Eighty-two Long Evans rats were divided into 6 groups and received acute or chronic IOP elevation to the right eye. Acute IOP elevation was induced by anterior chamber perfusion for 60min at 15mmHg (sham, n=10), 40mmHg (mild, n=12) or 130mmHg (severe, n=12). Chronic IOP elevation was induced by intracameral injection of an optically clear crosslinking hydrogel. A subset of animals (treated, n=19) received daily oral citicoline treatment (500mg/kg) for 7 days prior to, and every 48 hours for 14 days after gel injection, while another 19 rats were not treated orally (untreated). Ten rats received intracameral injection of buffer solution as sham control. IOP and visual acuity (VA) were measured by tonometry and optokinetics, respectively, for 5 weeks. Restingstate FC MRI was acquired at 9.4 Tesla covering the visual brain nuclei at 5 weeks after IOP elevation induction.
Result(s): Hydrogel injection significantly increased IOP for 5 weeks, with no difference between citicoline-treated (25.8+/-0.4mmHg) and untreated (26.4+/-0.5mmHg) groups (p>0.05). Severe acute IOP elevation resulted in lower VA [0.21+/-0.03cycles/degree (c/d)] than mild acute IOP elevation (0.38+/-0.03c/d) at week 5 (p<0.05), whereas VA of the citicoline-treated chronic group (0.31+/-0.01c/d) was higher than the untreated group (0.26+/-0.01c/d) (p<0.05). Larger magnitude of acute IOP elevation caused larger FC differences in subcortical visual nuclei (Fig. 1). In the chronic IOP elevation groups, untreated animals showed significantly decreased cortico-collicular, cortico-geniculate and subcortical FC, while citicoline treatment appeared to preserve or improve such FC (Fig. 2).
Conclusion(s): Increasing magnitude and duration of IOP elevation caused differential effects on visuomotor behavior and functional brain integrity, whereas citicoline ameliorated the effects. These findings merit further investigation into treating glaucoma beyond lowering IOP. (Figure presented)

Purpose: The purpose of this study is to evaluate medical student perception of the current ophthalmology curriculum without mandatory rotation at New York University School of Medicine (NYUSOM). Despite the lack of emphasis on ophthalmology in many medical school curricula, eye examination and management skills are important for physicians to master because they can reveal systemic pathology and require emergent treatment. In the context of rapidly evolving medical school curricula and lack of national ophthalmology education standards, it is important to assess ophthalmology training adequacy.
Method(s): A cross-sectional Internet survey was distributed to all currently enrolled NYUSOM students, including those pursuing dual degrees, in March to May 2017. The main parameters measured in the study were students' self-reported confidence with ophthalmology skills and satisfaction with curriculum.
Result(s): Response rate was 27.5% (166 of 604) of NYUSOM students. Many students reported they were not comfortable diagnosing eye emergencies (64%), using a direct ophthalmoscope (71%), or testing visual acuity (50%). The majority of students did not want ophthalmology to become a mandatory rotation, but reported additional in-person training would be most helpful, compared to videos, web-based didactics, lectures, or virtual training. Completion of an ophthalmology elective and more hours of ophthalmology training were associated with increased confidence with eye examination and greater satisfaction with the curriculum.
Conclusion(s): It is critical for all physicians-in-training to have adequate skills in eye examination. Identifying areas of improvement and determining the best teaching modality will be important in updating the ophthalmology curriculum for medical students. The majority of medical students are not at all or only slightly confident with eye examinations. Increasing the amount of in-person ophthalmology training in medical school improves confidence with eye examination. (Figure presented)

Purpose: Mentorship during various stages of medical training has been demonstrated to improve satisfaction in training and also to shape career goals. There are few studies evaluating formal mentorship programs within ophthalmology residency. We aim to evaluate the mentorship needs of ophthalmology residents, which may provide the framework to establish a formal mentorship program.
Method(s): An online cross-sectional survey was distributed in May 2017 to all residents (n=20) in the New York University School of Medicine ophthalmology residency program to evaluate their perspectives on mentorship. The questionnaire consisted of multiple choice and Likert-type questions.
Result(s): The response rate was 100% (20/20 of surveyed residents), of which 7 were PGY- 2, 6 were PGY-3, and 7 were PGY-4. Seventy-five percent (15/20) of residents reported that mentorship was "very important" during residency. Approximately two-thirds of residents (13/20) had participated in a formal mentorship program prior to residency. Eighty percent (16/20) of residents reported that two mentors were an ideal number, while 20% preferred three mentors. Sixty percent (12/20) of residents had already identified an informal mentor during residency. Respondents replied that "accessibility" was the most important quality in a mentor followed by "willingness to write a letter of recommendation". Gender, age, and academic ranking were the least frequently selected as important qualities in a mentor. The most common reason for pursuing a mentor was "seeking career guidance", followed by "obtaining a letter of recommendation" and "seeking someone as an advocate or confidant". The least commonly selected reason for seeking a mentor was "improvement of clinical skills".
Conclusion(s): Ophthalmology residents view mentorship as an important part of their training. Residents prioritize accessibility and career guidance as important aspects of mentorship and many are seeking a faculty member who may contribute a reference letter in the future. The results of this survey have contributed to the development of a formalized residency mentorship program, and help guide mentorship objectives and practice

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

In 400BC, Hippocrates wrote the first record of glaucoma. Since then, increasingly objective diagnostic techniques have enabled earlier detection of glaucoma and its progression, providing greater certainty in decision-making and early medical and surgical intervention.