This project will continue the development of a class of statistical methods for morphometrics, the quantitative analysis of biological shape. These methods apply to landmark data, locations of biologically corresponding (homologous) points in all the forms of a series or study. Such data are routinely collected for craniofacial biology, in many modes of diagnostic medical image analysis, and in evolutionary and developmental studies of whole organs or organisms. Recent advances in morphometrics have made possible the direct analysis of mean changes and other effects upon form and form-change when the forms are archived by lists of landmarks in this way. The description of any such contract or trend includes both a roster of distances and ratios most effectively capturing the phenomenon under study and statistical tests for various aspects of this description. The work proposed will extend this new morphometric praxis to include an additional source of information and to encompass studies of variation as well as of mean tendencies. (1) Biological outline data include information about the curving of form in-between landmarks, such as the shape of the human cranium or facial profile or the curving of the heart wall around the left ventricle. This information will be incorporated to extend the landmark-based statistical machinery. (2) At present, one can describe only one deformation, or one statistical mean deformation, at a time. The morphometric statistical algebra will be extended to extract interpretable components of deformation-homogeneous strains, spatial gradients, local distortions; to verify their reliability in the face of sampling variation; and to identify the features of particular deformations (e.g., instances of syndromes) as typical or atypical of a sample of deformations in one or more of these aspects. These new morphometric techniques will be applied to demonstration data sets from studies in craniofacial growth, craniofacial anomalies, and myocardial infarction. They are intended to apply more generally to greatly increase the efficiency with which information may be exploited in any biological or biomedical study of the large-scale shapes of organs or organisms, their variability, their changes over time, or their relation to standards of form as compiled in atlases.
Showing the most recent 10 out of 13 publications
