Segmentation of lateral ventricles in Magnetic Resonance Images

AGUSTINA BOUCHET; JUAN IGNACIO PASTORE; MARCEL BRUN; VIRGINIA BALLARIN

Paraná

Congreso; Congreso. VI Congreso Latinoamericano de Ingeniería Biomédica (CLAIB 2014); 2014

Consejo Regional de Ingeniería Biomédica para América Latina (CORAL)/Universidad Nacional de Entre Ríos

Magnetic Resonance Imaging is a powerful tool for the early diagnosis of different neurologic diseases. Given that this kind of images have high noise levels and imprecision in the definition of their structures, in this work we propose the use of Compensatory Fuzzy Mathematical Morphology to improve the segmentation of lateral ventricles, compared to Mathematical Morphology. To compare their performance we implement a classic Mathematical Morphology and a Compensatory Fuzzy Mathematical Morphology version of a segmentation algorithm. Once designed the algorithm, we evaluate the results obtained from the use of each methodology. Their performance is studied on a 160 images set, containing goldstandard images, provided by the FLENI institute, displaying a correct detection of 94.71% for Mathematical Morphology and 97.46% for Compensatory Fuzzy Mathematical Morphology.