Integrated fuzzy clustering

Cluster analysis is a valuable tool for exploratory pattern analysis, especially when the information available is not complete, and/or the data model is affected by uncertainty, and imprecision. Here a new integrated clustering method, based on a fuzzy approach is proposed. Its implementation consists of the combination of hierarchical fuzzy algorithms. The performance and accuracy of the methodology are tested on biomedical images. An application to the segmentation of magnetic resonance images is discussed.