Applying Geometric Morphometrics to Digital Reconstruction and Anatomical Investigation

Virtual imaging, image manipulation and morphometric methods are increasingly used in medicine and the natural sciences. Virtual imaging hardware and image manipulation software allows us to readily visualise, explore, alter, repair and study digital objects. This suite of equipment and tools combined with statistical tools for the study of form variation and covariation using Procrustes based analyses of landmark coordinates, geometric morphometrics, makes possible a wide range of studies of human variation pertinent to biomedicine. These tools for imaging, quantifying and analysing form have already led to new insights into organismal growth, development and evolution and offer exciting prospects in future biomedical applications. This chapter presents a review of commonly used methods for digital acquisition, extraction and landmarking of anatomical structures and of the common geometric morphometric statistical methods applied to investigate them: generalised Procrustes analysis to derive shape variables, principal component analysis to examine patterns of variation, multivariate regression to examine how form is influenced by meaningful factors and partial least squares analysis to examine associations among structures or between these and other interesting variables. An example study of human facial and maxillary sinus ontogeny illustrates these approaches.