Biological shape differences often are represented as diffeo-morphisms of a Cartesian coordinate grid. This paper suggests that their spatially discrete, localized features, for instance the details that suggest underlying developmental or pathological processes, can often be identified with variants of the singularity (x, y)-> (x, x2y + y3). This is an unfamiliar singularity, generic of codimension 1, at which a pair of cusps appears as a function of a parameter for “extrapolation.” I introduce canonical coordinates for such singularities and show how they may be used to produce objective reports of grids encountered in an empirical context. An example is shown involving the corpus callosum in Fetal Alcohol Syndrome. These features appear to be robust under relaxation of bending energy against Euclidean distance, the analogue to multiscale analysis for discrete punctate data.
CITATION STYLE
Bookstein, F. L. (1998). Singularities as features of deformation grids. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1496, pp. 788–797). Springer Verlag. https://doi.org/10.1007/bfb0056266
Mendeley helps you to discover research relevant for your work.