Artefact reduction on CT images of fossils to allow 3D visualisation

The evaluation of non-medical subjects with computed tomography (CT) is widespread and has been applied to the study of fossils. This is particularly useful when fossils are fully or partially encased in host rock, the removal of which would prevent a more detailed analysis of the post-death history of the specimen (it's taphonomy). While a series of CT images can be segmented and rendered to give 3D visualisation, the occurrence of CT artefact can seriously reduce the quality of such images. The current project investigated a method to reduce such artefacts allowing good 3D fossil visualisation. The artefact of most concern is beam hardening, which is particularly evident when the attenuation along projection rays varies greatly. Such a case occurs when discreet high attenuation materials are found in generally low attenuation material, but particularly when materials have irregular shape. The artefact reduction method used involved encasing the specimen prior to CT, to form a cylindrical shape using a manufactured phantom material of similar attenuation. The fabrication of phantom materials to simulate water or biological materials is well established and is achieved with materials using epoxy and gelatine bases amongst others. The use of a gelatine base was selected, as the specimens had to be recovered after investigation. Software was developed to calculate the effective photon energy of the CT scanner in the required material. This was then used, with the 'mixture rule', to iteratively determine the composition of the phantom material. As the attenuation of fossils and their host is significantly higher than biological materials, high Z fillers were used to get the required attenuation. This reduced the energy range for which the materials were equivalent. Fabrication proved challenging however the encasement of angular rock samples with phantom material demonstrated that phantom materials were fabricated with the required attenuation and that artefact occurrence was effectively eliminated. In many cases the rock shape does not generate significant artefact. In other cases the use of water may assist with artefact removal. For some small samples the above process is necessary and gives good artefact reduction. For larger samples non-gelatine based materials my need to be investigated. (author)