The 84th Annual Meeting of the American Association of Physical Anthropologists (2015)


Statistical evaluation of cortical thickness maps of the humeral diaphysis: A comparison of techniques

THOMAS G. O'MAHONEY, WILLIAM I. SELLERS and ANDREW T. CHAMBERLAIN.

Faculty of Life Sciences, University of Manchester

March 27, 2015 , Archview Ballroom Add to calendar

The aim of cortical thickness mapping is to visualise spatial variation in the thickness of cortical bone throughout a specimen. The technique also provides a means of statistically comparing the cortical properties of different individuals. Cortical mapping is becoming more prevalent in the literature as an aid in assessing morphological differences in long bones that may be ascribed to habitual behavioural differences or to genetic divergence. It has also been used in the medical literature to assess bone loss through osteoporosis.

At present, four forms of size standardisation and superimposition of the maps are popular for inter-individual comparisons: fourier transform; procrustes superposition; statistical parametric mapping and iterative closest point. All apart from the last operate directly on the raw measurements, whereas ICP relies upon a pre-generated 3d mesh of the periosteal and endosteal surfaces, which are deformed to a standard model. After superposition the residual variation between individuals can be analysed using PCA or Discriminant Functions.

Cortical thickness maps were generated for the 20-80% of length margins of the diaphysis from a large sample (N>100) of juvenile and mature modern human humeri and analysed using all four techniques. We establish that one can potentially track the development of muscular attachments along the bone, using this technique. We discuss how using the different standardisation processes can alter our conclusions as to group affinities, with implications for methodological selection when examining fossil remains.