Multiresolution based fuzzy c-means clustering for brain hemorrhage analysis

An automatic image segmentation technique is developed for segmenting the hematoma area from brain CT images. The features of an image are firstly extracted based upon the multiresolution method, and then fuzzy c-means clustering technique is applied for optimal classification. It is compared to other thresholding method such as fuzzy c-means, competitive Hopfield neural network, and fuzzy Hopfield neural network. From the results, it is shown that this proposed method is superior to those thresholding techniques. It is very useful and helpful for the physicians in studying the relationship between the size of hematoma and the clinical symptoms.