M. Salas Zuniga, "Extracting Skull-Face Models from MRI datasets for use in Craniofacial Reconstruction", PhD thesis, Department of Computer Science, The University of Sheffield, 2010. (Supervisor: Dr Steve Maddock). (pdf 9Mb)
PhD Abstract: This thesis addresses the problem of how to improve the computer-based methods for forensic craniofacial reconstruction. The aim of forensic craniofacial reconstruction techniques is to estimate the face of a person when only the skull remains are available for identification. From manual to computerised techniques, different solutions have been proposed for estimating the face from the skull in a reliable and consistent way. In both types of systems, an issue is that the results produced are not reproducible consistently and are difficult to evaluate. This issue arises mainly because the reconstruction methods are based on a limited type of anthropometric data about the skull and face relationship. In this thesis, a method to improve the craniofacial reconstruction process is developed by focusing on two aspects. First, a method is developed which uses MRI data to produce detailed data to analyse the skull and face relation. 3D skull and face models can be generated in a systematic way, offering the possibility to extend these sources as needed (i.e. it is possible to create data sources designed for specific populations). Second, a craniofacial reconstruction system is created that exploits the advantages of the detailed 3D skull and face models. As a result, the system can produce estimations of faces that can be assessed quantitatively, and produced in a systematic way.