Application of Digital Image Processing Techniques for Segmentation of Meningiomas in Magnetic Resonance Images

Brain tumor is of the main diseases that affect the brain, meningioma being the most common among all types of brain tumors. This pathology has a benign character in most situations. In order to aid in a rapid and non-invasive prognosis, magnetic resonance imaging is used as the means of diagnosis. Therefore, with the objective of developing a tool to assist in the analysis of the lesion and medical decisions, an algorithmic strategy of automatic segmentation was developed for magnetic resonance images of meningiomas, in order to delimit the tumor region from the application of processing techniques digital images. The results obtained in this analysis corresponded to about 66% of correct answers in the segmentation of the region of interest.