• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.18.2013.tde-27032013-092631
Document
Author
Full name
Renan Caldeira Menechelli
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2013
Supervisor
Committee
Schiabel, Homero (President)
Martinelli, Simone Elias
Traina, Agma Juci Machado
Title in Portuguese
Caracterização de sinais secundários em imagens mamográficas por redes neurais artificiais para auxílio ao diagnóstico do câncer de mama
Keywords in Portuguese
Achados secundários
Assimetria tecidual
Câncer de mama
Processamento de imagens
Reconhecimento de padrões
Abstract in Portuguese
O constante aumento do número de novos casos de câncer de mama vem despertando interesse na elaboração de módulos de esquemas CAD a fim de proporcionar um diagnóstico de maior precisão. Entretanto, a maioria das pesquisas está empenhada em detectar ou classificar fatores primários presentes em imagens mamográficas, como módulos e microcalcificações. Áreas assimétricas, retração de mamilo, linfonodos axilares, entre outros, são considerados como fatores secundários no diagnóstico do câncer de mama, apesar de poderem alertar para o surgimento não só dessa, mas de outras doenças no futuro. Por isso, essa pesquisa contempla a implementação de um sistema computacional capaz de auxiliar na detecção e classificação, conforme padrão BI-RADS®, de regiões que contenham sinais secundários capazes de levantar suspeitas da presença ou surgimento do câncer de mama, em imagens mamográficas digitais, utilizando técnicas inteligentes e automáticas de processamento de imagens e redes neurais artificiais. A acurácia alcançada em cada etapa foi: detecção de assimetria de 82,8%, retração de mamilo de 95% e Az = 0,93, detecção de linfonodos axilares = 74,9%. Objetiva-se que o resultado do trabalho seja inserido como um dos módulos de um protótipo de esquema CADx em mamografia, a fim de ampliar o conjunto de informações a serem usadas na classificação de cada caso sob análise, visando o aumento da precisão diagnóstica.
Title in English
Characterization of secondary signals in mammographic images by artificial neural networks to aid diagnosis of breast cancer
Keywords in English
Image processing
Mammography asymmetry
Nipple retraction
Pattern recognition
Secondary findings
Abstract in English
The increase in the number of cases of breast cancer have attracted interest in developing modules of CAD schemes to provider higher diagnostic accuracy. However, most researches are engaged in detect and classify primary factors present in mammographic images such as nodules and microcalcifications. Asymmetric areas, nipple retraction, axilary limph nodes, among other, are considered as secondary factors to diagnostic the breast cancer, although they may alert for the emergence not only of this but of other diseases in the future. Thus, this research includes the implementation of a computer system able to assist in the detection and classification, according to BI-RADS®, of regions that containing secondary signals able to arousing suspicion of the presence or appearance of breast cancer in digital mammographic images using intelligent and automatic techniques in the image processing and artificial neural networks. The accuracy obtained in each step was: detection of asymmetry of 82.8%, nipple retraction of 95% and Az = 0.93, detection of axilary lymph nodes = 74.9%. The purpose is that the result of the work is entered as one of the modules of a prototype of CADx schem in mammography in order to extend the range of information to be used in the classification of each case under analysis, aiming to increase diagnostic accuracy.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Renan.pdf (12.48 Mbytes)
Publishing Date
2013-03-28
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.