Statistical shape analysis: From landmarks to diffeomorphisms
The study of anatomical shape changes and of their relationship
with disease processes is one of the central topics in medical image
analysis. For example, many neurodegenerative disorders cause
widespread brain shape changes that can be observed in 3D brain
MRI scans, as illustrated in Fig. 1. Identifying statistical shape differences
between healthy subjects and patients affected by a disease
promises to provide new clinical insights and ultimately improve
diagnosis and treatment. In this paper, we briefly review the
evolution of representations and statistical modeling techniques
used for anatomical shape analysis and highlight recent developments
that take us closer to routine applications of shape analysis
based on large collections of medical images.