Slice, D.E. 2007.Geometric Morphometrics. Annual Review of Anthropology 36: 261-281.
(Volume publication date September 2007) ABSTRACT. Morphometrics, the field of biological shape analysis, has undergone major change in recent years. Most of this change has been due to the development and adoption of methods to analyze the Cartesian coordinates of anatomical landmarks. These geometric morphometric (GM) methods focus on the retention of geometric information throughout a study and provide efficient, statistically powerful analyses that can readily relate abstract, multivariate results to the physical structure of the original specimens. Physical anthropology has played a central role in both the development and the early adoption of these methods, just as it has done in the realm of general statistics, where it has served as a major motivating and contributing force behind much innovation. This review surveys the current state of GM, the role of anthropologists in its development, recent applications of GM in physical anthropology, and GM-based methods newly introduced to, or by, anthropology, which are likely to impact future research.
Acronyms and Definitions
Bending energy: minimized in the construction of thin-plate splines; measures the bendedness or kinkiness of the interpolated space between landmarks
Geometric morphometrics (GM): morphometric methods, often based on Cartesian coordinates of landmarks, that retain all geometric information in the data throughout an analysis
GPA: generalized Procrustes analysis
Morphometrics: examines central tendencies of shape, shape variation, group differences in shape, and associations of shape with extrinsic factors
PCA: principal components analysis
Procrustes analysis: the analysis of shape coordinates generated by the least-squares superimposition of configurations of landmarks
Shape: geometric properties of an object invariant to position, orientation, and isometric (global magnification/reduction) size differences
Shape variable: geometric variables invariant to position, orientation, and isometric size. Superimposed landmark coordinates are shape variables. Raw coordinates are not
Thin-plate spline: a visualization tool based on the interpolation of landmark differences mapping one set of landmark locations exactly onto another
No hay comentarios:
Publicar un comentario