Multimodal (MRI/PET) Fusion of Cerebral Images in Multiresolution Domain

As a result of the development of imaging techniques and information technology, it is possible to have several heterogeneous data. The fusion an anatomic imaging such as the magnetic resonance imagery (MRI) and functional imaging type such as the positron emission tomography (PET), made it possible to lead to a greater complementarities of data. In this work, we present a comparative study of fusion techniques in the transformed domain of the images. The results obtained showed that the best combination is the simultaneous use of the method of fusion DWT and the technique of classification K-Means. Keywords—Fusion of images; MRI; PET; cerebral images; multiresolution; pyramidal decomposition; segmentation; classification; k-means, Fuzzy C-Means; Maximisation Estimate.