MLMI Praktikum Report - Multi Modal Similarity

Different imaging modalities show different information. In medical diagnosis as well as in many image recognition tasks, we need to align the images to have all the information concentrated at each point. The bone of contention is to define a similarity measure for aligning images. Out of the many approaches and algorithms to measure similarity of objects, in this report, we intend to demonstrate two approaches for similarity measurement of 2-D brain images, namely clustering by fast search and find of density peaks and forests of extremely randomized binary trees.