Sistema de detección asistido por ordenador para ictus

This paper presents a system for detection of ischemic and haemorrhagic stroke in computed tomography images. Firstly, the original image is converted into grey scale and the noise is removed using the bilateral filter. Then skull regions are removed by a morphological function. The image is classified into infarct or haemorrhagic stroke based on the mean and standard deviation. A method for manual segmentation was also implemented. This method is based on the use of deformable models, specifically applying the gradient vector flow Snake algorithm. The results are segmented images for abnormal region. Experiments carried out on computed tomography images show the accuracy of the proposed system.