A New Landmark-Independent Tool for Quantifying and Characterizing Morphologic Variation

This paper develops a landmark-independent, deformable-registration-based framework that can utilize 3D surface images generated by any multidimensional imaging modality. The framework provides compact representations of image differences that are used to assess and compare potentially biologically relevant changes in 3D shape. The utility and sensitivity of the tools developed in this work are demonstrated using similarity retrieval of shape changes in a normal developmental time series of chick embryos. The results motivate future use of these tools for defining trajectories of normal growth, aiding research into conditions causing disruptions to normal growth.