Search

Background: To determine the diagnostic accuracy of panoramic radiograph and magnetic resonance imaging (MRI) for detection of signs of temporomandibular joint (TMJ) degenerative joint disease (DJD). Methods: Panoramic radiographs, bilateral TMJ MRI and bilateral TMJ computed tomography (CT) were performed on 705 subjects. Three calibrated board-certified radiologists blinded to the subjects’ clinical findings interpreted all images. Assessment of diagnostic accuracy of panoramic radiographs and MRI for detecting signs of DJD was compared to the reference-standard diagnoses derived from the CTs. DJD was defined by the presence of subcortical cyst, surface erosion, osteophyte or generalized sclerosis. Target sensitivity and specificity were > 70% and > 95%, respectively. Results: For panoramic radiographs, sensitivity and specificity were: subcortical cysts - 14%, 100% respectively; erosion - 20%, 100% respectively; osteophyte - 12%, 100% respectively and generalized sclerosis - 33%, 100%, respectively. For MRI, sensitivity and specificity were: subcortical cysts - 32%, 100% respectively; erosion - 35%, 99% respectively; osteophyte 71%, 98% respectively and generalized sclerosis 50%, 100% respectively. For diagnosis of signs of DJD based on panoramic radiographs, radiologists’ inter-examiner reliability was slight (k=0.16), moderate (k=0.47) when using MRI and substantial with CT images (k=0.71). Conclusions: Panoramic radiographs and MRI had below target sensitivity but above target specificity in detecting all CT-depicted signs of DJD with the exception of detection of MRI-depicted osteophytes, which had adequate diagnostic accuracy. Practical Implications: Use of CT to diagnosis signs of TMJ DJD is recommended to address the false negatives that can occur with panoramic radiographs and MRI.

Over the course of the current decade, microRNA-mediated regulation has been recognized as a major mechanism controlling gene expression at the post-transcriptional level. miRNAs regulate numerous and diverse aspects of biology, including development, differentiation, proliferation, metabolism and immunity. A single miRNA can regulate hundreds of genes and a gene can be regulated my multiple miRNAs. miRNAs suppress gene expression by binding to mature mRNA transcripts and promoting mRNA degradation, inhibiting translation or both. Identifying “true” targets still remains a challenge due to lack of perfect complementarity of binding of the miRNA to target mRNAs, resulting in the prediction of thousands of targets. In this dissertation, we seek to use bioinformatics approaches to narrow down these target predictions and place them in specific biological contexts. We begin by building a gene annotation enrichment and visualization tool that can be used to characterize the functions of miRNA targets with respect to a number of functional features. We then use protein-protein interaction networks as a basis for miRNA target prioritization and identification of potential miRNA-regulated modules. We find that many of the central proteins in protein-protein interaction sub-networks of miRNA targets are known targets. We also find that modularity within a miRNA-regulated interactome could be biologically meaningful and could serve as a way to dissect the diversity of functions regulated by a miRNA. Looking at another aspect, we integrate miRNA and gene expression analyses to narrow down the target predictions. Using expression data from two diseases, we identify facets of miRNA involvement, some that are already known and others that are novel. The results found here show the feasibility of our approaches to identify potentially important targets of miRNAs. Further extension and generalization of these approaches could equip researchers with useful tools for sifting through the considerable amounts of existing miRNA targets information.

This thesis covers the lessons I've learned throughout film school and how I have applied them towards the making of my thesis film, the 12-minute narrative short, Light, shot on digital video. Every aspect of the filmmaking process is covered including my education at UNO and its application to my thesis. Areas covered include the writing process, pre-production, production, and post-production. The one area that is omitted is distribution, something that was not taught at UNO and something that I am in the process of doing. Keywords: Film Production MFA Short Film India Kerala

In recent years, molecular imaging, defined as the visualization of biologic process in vivoat the molecular level, has gained prominence in the detection and monitoring of cancerous tissue. Several targets for the molecular imaging of cancer have been identified but in order to produce selective contrast enhancement, target specific contrast agent delivery methods and optimized imaging techniques need to be developed. This dissertation describes a molecular imaging approach that uses liposomes loaded with Ultra-small Super-Paramagnetic Iron Oxide (USPIO) nanoparticles targeted to tumor cells. Liposomes made from dioleylphosphatidylserine, (DOPS) are being used as carriers for Saposin-C, a fusogenic protein, currently being investigated for its effect in inducing cell death in a variety of human cancer cells. We developed an efficient method for loading Saposin-C-DOPS liposomes with USPIO. Further, we tested the uptake and MR detectability of these contrast-laden liposomes, referred to as ScDOPS-IO, in vitroin tumor cell cultures. Finally, we evaluated the feasibility of detecting ScDOPS-IO liposomes using MR imaging in vivo. The ScDOPS-IO liposomes encapsulated an average of 94.8±12.8 µg Fe /ml when a 1:30 molar ratio of saposin-C to DOPS was used with 1 mM DOPS and exhibited a transverse relaxivity (r2) of 101.21±4.9 mM-1s-1. Uptake in tumor cell cultures was found to be proportional to both time of exposure and initial concentration of ScDOPS-IO in growth medium. The cells were MR detectable against a uniform agarose background when average encapsulation was greater that 1 pg Fe/cell, using high resolution T2* weighted MR imaging. In vivoexperiments indicate accumulation of ScDOPS-IO in tumor xenografts in mice and T2 mapping of the xenografts show a significant drop in mean tumor T2 about 4 hours post injection of ScDOPS-IO. These results demonstrate the feasibility of using MR imaging for quantitative estimation of delivery and uptake of targeted drugs and early detection of tumors using targeted contrast agents. The methods presented here are promising for translation to applications in human subjects for clinical trials of new chemotherapy drugs.

Determining optical flow is a fundamental problem in computer vision and image analysis. Optical flow is the projection of three dimensional motion of objects onto a plane, estimated from a sequence of two dimensional images. However, as this is an ill posed problem, i.e. has no unique solution based on available information thus requiring additional constraints to make the problem well posed. We propose an algorithm based on feature matching using Hopfield Networks to determine the optical flow field between a sequence of images. We formulate the correspondence between features extracted from the two images as an optimization problem using a new robust disparity measure. The Hopfield network is then used to minimize the cost function and obtain an optimal match between the features in successive frames. This thesis offers a review of relevant literature and a detailed description of the algorithm. We then evaluate the performance of the algorithm on a variety of test sequences, both artificial and real. The results are found to be comparable to those obtained by other feature based algorithms, which suggests the use of Hopfield type networks in practical vision algorithms.