New Brain Extraction Method Using Expectation Maximization and Mathematical Morphology

The hi-tech development has led great success in both scientific and technical fields. Therefore we can obviously note that data processing takes less execution time and a smaller amount of human efforts. This is the advantage of the automatic line of work which aimed to reduce either physical or intellectual human effort. Furthermore it focused to let the painful and complex tasks to the machine including the computer. In our framework, the processing of medical images (brain MRI) is transferred to the computer so the setting up of processing algorithms becomes an axis of rich researches highly open to discussion and development in our work, and we know that MRI is an effective examination of disease diagnosis, In this paper, we present our own method of brain extraction inspired by different methods in the literature namely brain method (BET; BSE, MSPL; MCSTRIP; TMBE ...) and different segmentation algorithms. Then, we compare our own way to other methods with excellent results using supervised evaluation criteria both outline and region.