Model-based analysis of mammogramsReportar como inadecuado




Model-based analysis of mammograms - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Reference: Cerneaz, Nicholas J., (1994). Model-based analysis of mammograms. DPhil. University of Oxford.Citable link to this page:

 

Model-based analysis of mammograms

Abstract: Metastasised breast cancer kills. There is no known cure, there are no known preventativemeasures, there are no drugs available with proven capacity to abate its effects. Early identificationand excision of a malignancy prior to metastasis is the only method currently availablefor reducing the mortality due to breast disease.Automated analysis of mammograms has been proposed as a tool to aid radiologists detectbreast disease earlier and with greater efficiency and success. This thesis addresses some of themajor difficulties associated with the automated analysis of mammograms, in particular thedifficulties caused by the high-frequency, relatively insignificant curvi-linear structures (CLS)comprising the blood vessels, milk-ducts and fibrous tissues. Previous attempts at automationhave been overlooked these structures and the resultant complexity of that oversight has beenhandled inappropriately.We develop a model-based analysis of the CLS features, from the very anatomy of the breast,through mammography and digitisation to the image intensities. The model immediatelydictates an algorithm for extracting a high-level feature description of the CLS features. Thishigh-level feature description allows a systematic treatment of these image features prior tosearching for instances of breast disease.We demonstrate a procedure for implementing such prior treatment by 'removing' the CLSfeatures from the images. Furthermore, we develop a model of the expected appearance ofmammographic densities in the CLS-removed image, which leads directly to an algorithmfor their identification. Unfortunately the model also extracts many regions of the imagethat are not significant mammographic densities, and this therefore requires a subsequentsegmentation stage. Unlike previous attempts which apply neural networks to this task, andtherefore incorporate inherent insignificance as a consequence of insufficient data availabilitydescribing the significant mammographic densities, we illustrate the application of a newstatistical method (novelty analysis) for achieving a statistically significant segmentation ofthe mammographic densities from the plethora of candidates identified at the previous stage.We demonstrate the ability of the CLS feature description to identify instances of radial-scarin mammograms, and note the suitability of the CLS and density descriptions for assessmentof bilateral and temporal asymmetry. Some additional potential applications of these featuredescriptions in arenas other than mammogram analysis are also noted.

Type of Award:DPhil Level of Award:Doctoral Awarding Institution: University of Oxford Notes:This thesis was digitised thanks to the generosity of Dr Leonard Polonsky

Contributors

Brady, MichaelMore by this contributor

RoleSupervisor

 

Michael BradyMore by this contributor

RoleSupervisor

 Bibliographic Details

Issue Date: 1994Identifiers

Urn: uuid:a8d91bb2-429c-4da3-9f1b-6209771c61b5

Source identifier: 603835692 Item Description

Type: Thesis;

Language: eng Subjects: Breast Radiography Metastasis Tiny URL: td:603835692

Relationships





Autor: Cerneaz, Nicholas J. - institutionUniversity of Oxford facultyMathematical and Physical Sciences Division - - - - Contributors Br

Fuente: https://ora.ox.ac.uk/objects/uuid:a8d91bb2-429c-4da3-9f1b-6209771c61b5



DESCARGAR PDF




Documentos relacionados