ABSTRACT
In last few decades Medical Image Processing became one of the most popular and challenging research field due to its non-invasive nature for viewing internal body organs. Medical imaging modalities such as MRI, CT scan mostly depend on computer imaging technology to generate or display digital images of the internal organs of the human body which helps the medical practitioners to visualize the inner portions of the body. MRI of brain provides best resolution images for visualizing brain tissue without any surgical interventions. We propose a multimodal histogram thresholding method to separate brain MR images into different tissue components and also dissociation of skull from brain. We are processing the histogram of MRI brain in such a way that, number of cluster and cluster center are automatically detected and we are passing this data to a clustering algorithm. In this technique no prior knowledge about the number of cluster is needed. The coordinates of cluster centers and their number can be passed to the clustering algorithm. In our work we used Fuzzy C-Means (FCM) for the clustering the gray levels. Each cluster gives us different gray levels for each tissue type in the MR brain image. Main Objective of our work is to dissociate the skull from a brain MRI and automatically segment different parts of brain tissue types e.g. white matter (WM), gray matter (GM), and Cerebra-Spinal Fluid (CSF).
