Computing metrics on anatomical shapes in computational anatomy

Metric distances can be used to quantify the notion of close and far on anatomical shapes as represented in images. This is achieved by computing diffeomorphic transformations between given images and measuring their size. Transformations that are "far" from identity represent larger differences in shape and size than those "close" to identity. Such metrics may find possible clinical applications such as the detection and study of shape and size changes in various diseases that manifest in shape and size changes of anatomical organs.