Glaucoma is the second leading cause of irreversible blindness in the United States and is the leading cause of blindness in African Americans. Cupping or excavation of the optic nerve, which sends the visual signal from the photoreceptors in the eye to the brain, is a chief feature of glaucoma. A similar excavated appearance of the optic nerve is also the primary clinical sign of other congenital malformations of the eye including optic nerve head coloboma, optic pit, and morning glory disc anomaly collectively termed cavitary optic disc anomaly (CODA). Clinical similarities between CODA and glaucoma have suggested that these conditions may have overlapping pathophysiology. Although risk factors are known, such as the elevated intraocular pressure (IOP) observed in some glaucoma subjects, the biological pathways and molecular events that lead to excavation of the optic disc in glaucoma and in CODA are incompletely understood, which has hindered efforts to improve diagnosis and treatment of these diseases. Consequently, there is a critical need to clarify the biological mechanisms that lead to excavation of the optic nerve, which will lead to improvements in our understanding of these important disease processes. Because of their similar clinical phenotypes and the limited therapy geared at lowering IOP in glaucoma patients, our central hypothesis is that genes involved in Mendelian forms of CODA would also be involved in a subset of glaucoma cases and may provide insight into glaucomatous optic neuropathy. The purpose of my research project has been to identify and functionally characterize the gene that causes congenital autosomal dominant CODA in a multiplex family with 17 affected members. The gene that causes CODA was previously mapped to chromosome 12q14 and following screening of candidate genes within the region that did not yield any plausible coding sequence mutations, a triplication of a 6KB segment of DNA upstream of the matrix metalloproteinase 19 (MMP19) gene was subsequently identified using comparative genomic hybridization arrays and qPCR. This copy number variation (CNV) was present in all affected family members but absent in unaffected family members, a panel of 78 normal control subjects, and the Database of Genomic Variants. In a case-control study of singleton CODA subjects, CNVs were also detected; we detected the same 6KB triplication in 1 of 24 subjects screened. This subject was part of another 3-generation autosomal dominant CODA pedigree where affected members each have the same CNV identified in the larger CODA pedigree. A separate case-control study with 172 glaucoma cases (primary open angle glaucoma = 84, normal tension glaucoma = 88) was evaluated for MMP19 CNVs, however none were detected. Although our cohort of CODA patients is small limiting our ability to accurately determine the proportion of CODA caused by MMP19 mutations, our data indicates that the MMP19 CNV is not an isolated case and additional CODA subjects may have MMP19 defects. Because of the location of the CNV, we evaluated its effect on downstream gene expression with luciferase reporter gene assays. These assays revealed that the 6KB sequence spanned by the CNV in CODA subjects functioned as a transcriptional enhancer; in particular, a 773bp segment had a strong positive influence (8-fold higher) on downstream gene expression. MMP19, a largely understudied gene, was further characterized by expression studies in the optic nerve and retina. Using frozen sections from normal donor eyes, we demonstrated that MMP19 is predominantly localized to the optic nerve head in the lamina cribrosa region with moderate labeling in the postlaminar region, and weak labeling in the prelaminar region and retina. We also evaluated MMP19 expression in relation to the cell types that populate the optic nerve such as astrocytes and retinal ganglion cells. The pattern of expression is consistent with MMP19 being a secreted protein accumulating in the extracellular spaces and basement membranes of the optic nerve. Our studies have identified the first gene associated with CODA and future research is focused on recapitulating CODA phenotypes in animal models and assessing the mechanism of MMP19 involvement during development.

A fully integrated optical chem/bio sensor complete with integrated source, chemically sensitive waveguide, detector arrays, and associated signal processing electronics on a Si-CMOS chip is a challenging, but highly desirable goal. An evanescent-wave multimode interferometric sensing element is a sensitive method for sensing, which is easily integrated on Si-CMOS. This work is concerned with the design, analysis, and demonstration of a planar multimode interferometric chem/bio sensor that is compatible with the fabrication constraints of Si-CMOS. A 4000-micron-long interferometric that can be adapted for different agents by a particular sensing layer has been fabricated on silicon using silicon dioxide and silicon oxynitride. Hexaflouro-isopropanol substituted polynorbornene is the sensing layer. This sensor has also been fabricated on a Si-CMOS circuit with embedded photodetectors. A sensor on silicon was demonstrated with a minimum detectable index change of 2.0x10-6 using an accurate gas delivery system and a custom hermetic waveguide test chamber. A modal pattern analysis strategy has also been developed to extract the optimal SNR from the measured modal patterns. An understanding of the noise processes and spatial bandwidth effects has enabled an experimentally-based prediction of the index sensitivity of a fully integrated multimode chem/bio sensor on Si-CMOS at 9.2 x10-7. Theoretically, the sensitivity enhancement of high over low index sensing layers and transverse-magnetic over transverse electric modes is described. Also, the sensitivity enhancement of higher-order-transverse modes has been quantified. The wide-angle beam propagation method has been used to simulate the sensor. This simulation showed the relation between the modal pattern repetition period and sensor sensitivity. Further, the modal coupling properties of the multimode y-junction have been described. A second multimode y-junction has been designed to change the modal excitation under the SL, and thus the sensitivity. The chemo-optic response of the `substituted polynorbornene' polymer., hexaflouro-isopropanol substituted polynorbornene to methanol, water, iso-propanol, and benzene has been measured. Also, its thermo-optic response has been measured. Athermal interferometric chem/bio sensors have then been suggested.

This dissertation consists of two topics. One is a "Multi-pass Fiber Optic Surface Plasmon Resonance Sensor (SPR)" and the other is a "Nano-metallic Surface Plasmon Lens." Since both topics involved surface plasmon, the title of this dissertation is named "Surface plasmon coupled sensor and nanolens." For a multi-pass fiber optic SPR sensor, a fiber optic 4-pass SPR sensor coupled with a field-assist capability for detecting an extremely low concentration of charged particles is first demonstrated. The multipass feature increases the sensitivity by a factor equal to the number of passes. The field-assist feature forces charged particles/molecules to the SPR surface, increasing the sensitivity by an additional factor of about 100. Overall, the sensitivity exceeds the one-pass SPR device by a factor of about 400. A 10 pM concentration of 47 nm diameter polystyrene (PS) latex beads and 1 ?M concentration of salt dissolved in DI water were detected within a few seconds by the combined system. The equivalent index resolution for atomic size corresponding to ionized chlorine in salt is 10-8. This technique offers the potential for sensitive and fast detection of biomolecules in a solution. Secondly, a 44-pass fiber optic surface plasmon resonance (SPR) sensor coupled with a field-assist capability for measurement of refractive index change due to positive and negative ions is shown. The field-assist feature forces ions to the SPR surface, causing the SPR signal response to change which reflects a decrease or increase in refractive index depending on whether positive or negative ions are being attracted to the surface. This technique offers the potential for the sensitive detection of cations and anions in a solution. For a nano-metallic surface plasmon lens, we analyze the transmission of a normally incident plane wave through an Ag/dielectric layered concentric ring structure using finite difference time domain (FDTD) analysis. The dependency of the transmission efficiency on the refractive index in slit is studied. The numerical analysis indicates that the focusing beyond diffraction limit is found even at the extended focal length comparable to the distance of 7" from the exit plane using a circularly polarized coherent plane wave, ?=405 nm. Especially, compared to an Ag-only structure, the Ag/ LiNbO3 structure amplifies the transmission power by a factor of 6. Therefore, this Ag/dielectric layered lens has the potential for significantly higher resolution imaging and optical data storage.

published_or_final_version / Anatomy / Master / Master of Philosophy

Purpose: Retinal ganglion cell (RGC) loss is a measure of the progression of many visual disorders. It is important to identify RGCs with good specificity, so RGC numbers can be reliably analyzed. The purpose of this study was to analyze the effectiveness of two current RGC markers: Brn3a immunohistochemistry and the expression of Thy1-CFP in the Thy1-CFP transgenic mouse. Methods: Rhodamine-?-isothiocyanate (RITC) retrograde labeling, immunohistochemistry, wholemount retinal imaging, western blot, cross sectional analysis and cell densities in uninjured control animals and 3, 5, 7 and 14 days post-optic nerve crush (ONC) or transection (ONT) were tabulated. Results: Brn3a positive (Brn3a+) cell density was significantly less than RITC positive (RITC+) cell density in control mice. After ON injury, Brn3a+ cell density did not decrease at the same rate as RITC+ cell density. The density of RGCs that express Brn3a was significantly less than the individual Brn3a+ and RITC+ cell density at all experimental time points. Thy1-CFP positive (Thy1-CFP+) cell density was significantly less than RITC+ in control mice and significantly more than RITC+ cell density 14 days after ON injury. Thy1-CFP co-localized with ChAT positive (ChAT+) cells 7 days after ONT. Conclusion: Brn3a and Thy1-CFP are not reliable markers of RGCs. Retrograde labeling remains one of the most reliable methods of labeling RGCs in mice.

Thesis (Ph.D.)--University of Western Australia, 2007.

Title: Gradient boosted segmentation of retinal fundus images Author: Matúš Goliaš Department: Department of Software and Computer Science Education Supervisor: Doc. RNDr. Elena Šikudová PhD., Department of Software and Computer Science Education Abstract: Over the recent years, there has been an increase in the use of automatic methods in medical diagnosis. A significant number of publications have analysed eye disorders and diseases. One of the most severe eye conditions is glaucoma. It damages optic nerves and causes gradual loss of vision. An essential step towards a faster diagnosis of this disease is accurate segmentation of the optic disc and cup. This task is difficult due to many retinal defects, different image acquisition techniques, and artefacts caused by imaging devices. This thesis describes an iterative threshold-based algorithm for extraction of the optic disc. An objective function quantifying object similarity to the optic disc is defined to direct the iteration. Following that, we introduce a superpixel-based classification algorithm for extraction of the optic cup. We propose the use of gradient boosted decision trees which outperform random forest and support vector machine. In addition, we evaluate the proposed algorithms and their alternatives on a publicly available retinal fundus...

Thesis: M.S., Massachusetts Institute of Technology, Department of Psychology, 1973 / Vita. / Bibliography: leaves 67-73. / M.S. / M.S. Massachusetts Institute of Technology, Department of Psychology

No description available.

<p> This thesis deals with the limiting factors in the design of a long-haul fiber-optic communication system, and the techniques used to suppress their resulting impairments. These limiting factors include both linear and nonlinear effects, such as fiber chromatic dispersion and the Kerr nonlinearity, and the modulator-induced nonlinearity. </p> <p> In Chapter 3, the conditional probability density function (PDF) of the received elect rical signal given transmitted bit '1 '/'0' for a coherent fiber-optic transmission system based on binary phase shift keying (BPSK) is mathematically derived. Both amplified spontaneous emission (ASE) noise and fiber nonlinearity are taken into account . The results show that the conditional PDF of given bit '1' or '0' is asymmetric when intrachannel four-wave mixing (IFWM) is dominant, while it becomes nearly symmetric when the variance of ASE is much larger than that due to IFWM. The standard deviation of the received signal is calculated analytically. The system parameters, including optimum dispersion map and pre-compensation ratio, are optimized by analytically calculating variance of IFWM. Significant computation efforts can be saved using this approach as compared to full numerical simulations of the nonlinear Schrodinger equation, without losing much accuracy. </p> <p> In Chapter 4, an improved 4-f time-lens configuration is proposed. Fourier transform (FT) and inverse Fourier transform (IFT) can be realized using time lenses such that there is no need for time reversal at the end. A typical 4-f configuration consists of two 2-f systems and a temporal filter. The first 2-f system consisting of a time lens and two dispersive elements produces the Fourier transform (FT) of the input signal. The temporal filter modifies the spectrum. The next 2-f system produces the inverse Fourier transform (IFT). A wavelength division demultiplexer and a higher-order dispersion compensator based on 4-f configuration are numerical implemented. One of the advantages of the time-lens-based temporal filtering technique is that the transfer function of the temporal filter can be dynamically altered by changing the input voltage to the temporal filter (amplitude/phase modulator) and therefore, this technique could be used for dynamic switching and multiplexing in optical networks. </p> <p> In chapter 5, a direct-detection optical orthogonal frequency division multiplexing (DD-0-0FDM) is realized using time lenses. Typically, in OFDM systems, discrete Fourier transform (DFT) is used at the transmitter and inverse discrete Fourier transform (IDFT) is used at the receiver. In this chapter, it is proposed to use continuous Fourier transform (FT) and inverse Fourier transform (IFT) using time lenses that replace DFT and IDFT in the electrical domain. The third- and higher-order dispersive effects can be considerably reduced using the proposed DD-0-0FDM scheme. </p> <p> In Chapter 6, a coherent optical orthogonal frequency division multiplexing (OFDM) (C0-0-0FDM) scheme using time lenses is analyzed. The comparison of performance between the proposed scheme and the conventional optical OFDM scheme using fast Fourier transform (FFT) and inverse FFT in the electrical domain is made. Both the Mach-Zehnder modulator (MZM) induced and fiber induced nonlinearities are investigated. Results show that the time-lens-based C0-0-0FDM performs almost the same as the FFT-based C0-0-0FDM when the message signal launched to MZM is low so that MZM operates in the linear region. The nonlinearity of MZM degrades the performance of FFT-based C0-0-0FDM drastically when the power of message signal becomes sufficiently large, but only has negligible impact on the time-lens-based C0-0-0FDM. A periodical driving voltage has been proposed to set up the time lens such that the maximally required driving voltage level is kept low within the time frame. The advantages using the time-lens-based C0-0-0FDM are that (i) FT can be done in optical domain almost instantaneously, whereas the FFT in digital domain is slow and requires significant computational efforts, (ii) optical domain Fourier transform has a large bandwidth (~THz) and therefore, FT /IFT can be performed at a large symbol rate. </p> <p> In Chapter 7, the digital backward propagation (DBP) has been studied both in orthogonal frequency-division multiplexing ( OFDM) and single-carrier (SC) fiber-optic transmission systems. 16 quadrature amplitude modulation (QAM) is used for both systems with the bit rate of 100 Gbjs. The results show that OFDM and SC with Nyquist pulses (SC-Nyquist) have a superior performance as compared to SC with raisedcosine pulses (SC-NRZ) when the DBP is used. The impact of electrical filter bandwidth and nonlinear phase/amplitude noise has also been investigated. The performance of perfect-BP-based OFDM/SC initially improves when the electrical filter bandwidth increases at high signal-to-noise ratio (SNR). The comparison of the effects of nonlinear phase/amplitude noise among OFDM, SC-Nyquist and SC-NRZ systems is made and it is shown that SC-NRZ systems significantly suffer from the effects of nonlinear phase/amplitude noise, which explains the performance advantage of OFDM/SC-Nyquist over SC-NRZ when the DBP used. </p> / Thesis / Doctor of Philosophy (PhD)