V Congreso Latinoamericano de Ingeniería Biomédica (CLAIB2011)

An Algorithm for Automatic Segmentation of Spontaneous Cerebral Hemorrhages
Rodríguez Morales Roberto

##manager.scheduler.building##: Palacio de Convenciones de La Habana
##manager.scheduler.room##: Sala 12
Fecha: 2011-05-20 09:00  – 10:30
Última modificación: 2011-04-14 12:21


Computerized tomography images (CTI) are widely used in the diagnosis of stroke. In this paper, an algorithm for automatic segmentation of human spontaneous intra-cerebral brain hemorrhages (ICH) from computerized tomography images is proposed. The proposed strategy consists of five main steps: elimination of white matter, image enhancement, detection of region-of-interest via a reconstruction, and automatic segmentation. The method combines, essentially, a set of tools of mathematical morphology. To evaluate the proposed algorithm was implemented a graphical interface, in MATLAB and a dataset of 30 patients studies (360 image slices) was used. Obtained results were compared with the ground truth (a manual segmentation carried out by specialized medical personnel). The method provides 98% accuracy in detecting the ICH; and achieves an average precision of 96% at the slice level. The goal of this algorithm is later carry out automatic morphometrical analysis for helping the diagnostic and prognostic of specialists

Texto completo: PDF