Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) of the breast is employed for breast cancer screening in women with cumulative lifetime breast cancer risk of more than 20-25% (US and EU guideline). Other indications include, but are not limited to, preoperative staging, evaluation of women treated with neoadjuvant chemotherapy and problem solving in case of inconclusive findings from other modalities. Compared to mammography, which is the image modality commonly used in regular screening, breast DCE-MRI presents higher sensitivity, especially in women with dense breasts. However, the analysis of breast MRI requires interpretation of four-dimensional DCE data, as well as correlation to multi-parametric data from other MRI imaging sequences, and is therefore a time consuming task which can lead to potential observer errors.<\/p>\n
This thesis focuses on the study and development of automated image analysis techniques and computer-aided detection systems to improve the clinical workflow and aid radiologists for reading and interpreting MRI images.<\/p>\n
Automated tools for the identification and segmentation of the anatomical structures visible in breast MRI, such as pectoral muscle, breast and fibroglandular tissue, have been developed and a high agreement between automatic segmentations and manual annotations provided by experts has been achieved. These tools are essential in order to focus the detection of cancer in the breast area, quantify breast density, which is a known risk factor for the development of breast cancer, and to provide automatic linkage of findings to prior studies or images from other modalities of the same patient.<\/p>\n
Furthermore, a novel computer-aided detection (CAD) system to detect breast cancer in DCE-MRI has been proposed. This method is fully automatic and has the potential to support radiologists during the analysis of DCE-MRI data by automatically prompting suspicious areas. The CAD system has achieved high sensitivity with an acceptable number of false-positives findings and is able to detect lesions that were overlooked or misinterpreted by experts in a breast screening program with MRI.<\/p>\n
To summarize, in this thesis we have developed automated image analysis techniques for breast cancer assessment in DCE-MRI of the breast. We have demonstrated that the developed methods perform well and have the potential to be incorporated in clinical practice to facilitate reading and interpretation of DCE-MRI.<\/p>\n
<\/p>\n
____________<\/p>\n
<\/p>\n
Dynamic Contrast-Enhanced Magnetic Resonance Imaging<\/i> (DCE-MRI) de la mama s\u2019utilitza en programes de monitoritzaci\u00f3 per a la detecci\u00f3 preco\u00e7 del c\u00e0ncer de mama en pacients d\u2019alt risc. Aquesta modalitat d\u2019imatge tamb\u00e9 s\u2019utilitza per a altres finalitats com ara la preparaci\u00f3 preoperat\u00f2ria, l\u2019avaluaci\u00f3 de la resposta al tractament del tumor amb quimioter\u00e0pia o com a modalitat d\u2019imatge alternativa quan no es pot extreure un diagn\u00f2stic clar amb les imatges d\u2019altres modalitats. En comparaci\u00f3 amb la mamografia, que \u00e9s la modalitat d\u2019imatge utilitzada en programes regulars de cribratge, DCE-MRI t\u00e9 una major sensibilitat, especialment en dones amb mames denses. Malauradament, l’an\u00e0lisi d\u2019imatges de MRI de mama \u00e9s una tasca feixuga perqu\u00e8 requereix la interpretaci\u00f3 de dades din\u00e0miques en quatre dimensions, aix\u00ed com la correlaci\u00f3 de dades multi-param\u00e8triques que provenen d’altres seq\u00fc\u00e8ncies de MRI. Aquest fet pot ocasionar errors de lectura i diagn\u00f2stic de c\u00e0ncer en DCE-MRI.<\/p>\n
Per facilitar la lectura i la interpretaci\u00f3 d’imatges DCE-MRI, es necessiten sistemes d’ajuda assistits per ordinador per la\u00a0 detecci\u00f3 i avaluaci\u00f3 del c\u00e0ncer de mama. Amb tal finalitat, en aquesta tesi s’han estudiat i desenvolupat noves t\u00e8cniques inform\u00e0tiques per l’an\u00e0lisi automatitzat d’imatges de DCE-MRI.<\/p>\n
S\u2019han desenvolupat algorismes automatitzats per a la identificaci\u00f3 i la delineaci\u00f3 de les diferents estructures visibles en la imatge, com s\u00f3n el m\u00fascul pectoral, el pit i el teixit dens, \u00a0i s\u2019ha obtingut alta concordan\u00e7a entre les segmentacions autom\u00e0tiques i les anotacions realitzades per experts. Aquestes eines poden ser utilitzades en m\u00faltiples aplicacions: focalitzar la detecci\u00f3 del c\u00e0ncer en l\u2019\u00e0rea del pit, quantificar el teixit dens, que ha estat reconegut com un factor de risc per al desenvolupament del c\u00e0ncer de mama, i per proporcionar la correlaci\u00f3 autom\u00e0tica entre zones sospitoses en imatges pr\u00e8vies de DCE-MRI o altres imatges de diferent modalitat del mateix pacient.<\/p>\n
A m\u00e9s a m\u00e9s, s\u2019ha dissenyat, implementat i avaluat un nou sistema d\u2019ajuda de detecci\u00f3 de c\u00e0ncer de mama en DCE-MRI assistit per ordinador (Computer-Aided Detection<\/i> \u2013 CAD). Aquest m\u00e8tode \u00e9s totalment autom\u00e0tic i t\u00e9 com a finalitat donar suport als radi\u00f2legs durant l\u2019an\u00e0lisi de DCE-MRI mitjan\u00e7ant la indicaci\u00f3 autom\u00e0tica de zones sospitoses. En l\u2019avaluaci\u00f3 realitzada, el CAD ha obtingut una alta sensibilitat amb un nombre acceptable de falsos de positius i tamb\u00e9 ha estat capa\u00e7 de detectar lesions omeses o que van ser mal interpretades per experts en programes de cribratge de c\u00e0ncer de mama amb DCE-MRI.<\/p>\n
En resum, en aquesta tesi s’han desenvolupat t\u00e8cniques automatitzades d’an\u00e0lisi d’imatge per a l’avaluaci\u00f3 i la detecci\u00f3 de c\u00e0ncer de mama en imatges de DCE-MRI. S’ha demostrat que els m\u00e8todes proposats ofereixen un bon rendiment. La seva integraci\u00f3 en la pr\u00e0ctica cl\u00ednica podria facilitar i optimitzar la lectura i la interpretaci\u00f3 d\u2019imatges DCE-MRI.<\/p>\n","protected":false},"excerpt":{"rendered":"
Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) of the breast is employed for breast cancer screening in women with cumulative lifetime breast cancer risk of more than 20-25% (US and EU guideline). Other indications include, but are not limited to, preoperative staging, evaluation of women treated with neoadjuvant chemotherapy and problem solving in case of inconclusive… <\/p>\n","protected":false},"author":1,"featured_media":3349,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"neve_meta_sidebar":"","neve_meta_container":"","neve_meta_enable_content_width":"","neve_meta_content_width":0,"neve_meta_title_alignment":"","neve_meta_author_avatar":"","neve_post_elements_order":"","neve_meta_disable_header":"","neve_meta_disable_footer":"","neve_meta_disable_title":"","footnotes":""},"categories":[75,12,58],"tags":[],"class_list":["post-3346","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-medical-imaging-lab","category-news","category-scientific-results"],"acf":[],"yoast_head":"\n