| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.72 MB | Adobe PDF |
Advisor(s)
Abstract(s)
O cancro é uma das causas de morte mais frequente em todo o mudo e o cancro da mama é o mais comum em mulheres. Logo, é de extrema importância a deteção atempada, pois o diagnóstico precoce e o tratamento do cancro mamário aumentam significativamente as probabilidades de tratamento bem-sucedido.
O exame da mama através de imagens por ultrassom (US) é o principal coadjuvante da mamografia, tornando-se deste modo um exame ainda mais importante, complementado por sistemas CAD, detetando e classificando nódulos mamários.
Na presente dissertação foi desenvolvida uma metodologia, baseada em Operadores Morfológicos, Deteção de contornos, Métodos de Thresholding entre outros, para segmentar automaticamente imagens de US mamários, que determina diversos parâmetros com auxílio de Algoritmos Genéticos (AG).
Cancer is one of the most frequent causes of death all over the world and breast cancer is the most common in women. Therefore, it’s extremely important beforehand detection, since early diagnosis and treatment of breast cancer increases significantly the chances of successful treatment. The breast examination through ultrasound images (US) is the main assistant method to mammography, therefore becoming an even more important exam, complemented by CAD systems, detecting and classifying breast masses. In present dissertation it was developed a methodology, based on Morphological Operators, Contour Detection, Thresholding Methods among others, for automatically segment US breasts pictures, that determines several parameters with the support of Genetic Algorithms (GA).
Cancer is one of the most frequent causes of death all over the world and breast cancer is the most common in women. Therefore, it’s extremely important beforehand detection, since early diagnosis and treatment of breast cancer increases significantly the chances of successful treatment. The breast examination through ultrasound images (US) is the main assistant method to mammography, therefore becoming an even more important exam, complemented by CAD systems, detecting and classifying breast masses. In present dissertation it was developed a methodology, based on Morphological Operators, Contour Detection, Thresholding Methods among others, for automatically segment US breasts pictures, that determines several parameters with the support of Genetic Algorithms (GA).
Description
Keywords
Algoritmos Genéticos Cad Cancro da Mama Segmentação Ultrassom Mamário