PADIS

Pubblicazioni Aperte DIgitali Sapienza > Informatica e sistemistica "Antonio Ruberti" > INGEGNERIA DEI SISTEMI >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10805/1628

Title: Analisi Statistica Di Pseudo Immagini OCT Della Retina Per Scopi Diagnostici
Authors: RUSSO, VALENTINA
Tutor: Bruni, Carlo
Keywords: Diabetic Retinopathy
Optical Coherence Tomography
Stretched Exponential DIstribution
Statistical Analisys
Issue Date: 21-Apr-2011
Abstract: If you compare the eye to a camera, the retina is the sensitive film. The retina is a thin layer of nervous tissue at the level of which are focused the rays of light that pass through the dioptric media, produces an image that is transmitted via the optic nerve to the cerebral cortex where the image is interpreted. The retina is part of the brain and serves as a transducer to convert light energy into neuronal signals. It seems clear, therefore, like any disease that damages this membrane, can affect the whole functioning of the visual system. There are many diseases that can affect the retina, both direct and indirect, between the most common one may found: diabetic retinopathy, age-related macular degeneration and retinitis pigmentosa. While the latter seems to be caused by a mutation in the genes, the first two have a high incidence of pathology. Today, these diseases are diagnosed using a variety of devices that allow different techniques to obtain images of the retina such as to highlight any anomalies in it. In particular one of the most recent and promising applications of diagnostic imaging is Optical Coherence Tomography (OCT). This technique developed over the past 15 years provides benefits for both the patient and the quantity and quality of information they can provide. First of all is a non-invasive technique that does not require the use of contrast media, and allows, by exploiting the differential reflectivity of the various layers of the retina to the light beam incident on it, to scan the membrane depth with a resolution of the order of tens of mm. This work is part of a research project which aims to be able to detect, by a statistical analysis of data from the acquisition of OCT images, retinal abnormalities, whether or not related to diseases both directly and indirectly to the retina, in a very early stage as to be barely visible to the eye of specialists, allowing a diagnosis as early as possible. The retina has a strictly organized structure layers of cells in different specializations. The hypothesis of the study is playing is to be able to identify within each layer the same statistical behavior of the data (homogeneity), that is a definite probability distribution: the Strectched Exponential. The anomalies, caused by various conditions, appear on the retinal images as blots, i.e. more or less extensive areas characterized by a non-homogeneous statistical behavior.To identify these anomalies in this thesis have followed different approaches to statistical analysis. A first study was to analyze the variability of the OCT data within individual layers of the retina. It was noted that there are anomalies in the distribution of the data showed trends rather graphic variables (i.e. the presence of peaks in the graphs clearly marked on the distrubution of the parameters characterizing the spatial distrubution Stretched Exponential). A second approach has seen the use of Kolmogov-Smirnov statistical test for analysis of the parameters characterizing the Gaussian distribution. In this case, the analysis did not stop only at the identification of anomalies, but aggregating the data it was possible to determine a method that, through automatic analysis of data, can be supportive and provide diagnostic indication that will be deepened by the medical specialist. Also studying the distribution of the results of Kolmogorov-Smirnov test, for each individual patient, it was possible to determine a geometric distribution of the retinal spots: If a blot is present in a layer then it is likely that it can also be present in the adjacents and also a blot of considerable size can span multiple scan windows, showing thus a third dimension in the same retina.
URI: http://hdl.handle.net/10805/1628
Appears in PhD:INGEGNERIA DEI SISTEMI

Files in This Item:

File Description SizeFormat
RussoValentina-TesiDottorato-XXIII ciclo-DIS.pdftesi di dottorato ed articolo correlati28.71 MBAdobe PDF

File del Curriculum Vitae:

CurriculumVitae.pdf 227.85 kBAdobe PDF


This item is protected by original copyright

Recommend this item

Items in PADIS are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0! DSpace Software Copyright © 2002-2010  Duraspace - Feedback Sviluppo e manutenzione a cura del CINECA