Cortex Classification of the Infantile Brain in MRI Images Using Fuzzy Logic

Hypoxic ischemic encephalopathy (HIE) caused by asphyxia in the womb causes the decrement of the white matter (WM). Therefore, calculating the volume of the cerebral tissues for the infant with such symptom helps us for the purpose of quantifying the acuteness of symptom. Many methods for classifying the adult brain tissues with magnetic resonance (MR) images have been studied. However, these methods cannot be applied to classify the infantile brain tissues because the WM undergoes a myelination process in infantile brain, and the infantile brain image features are very different from adult one. This paper aims to propose a method for classifying the brain tissues in the myelination process. The proposed method addresses the intensity nonuniformity (INU) artifact by locally adapting a fuzzy spatial model of MR signals. The fuzzy model represents transition of MR signals on a line from the cerebral contour to inside the cerebrum. By using the fuzzy spatial model, the proposed method assigns fuzzy degree belonging to the cerebral cortex into voxels dependent of their locations.