Computed Tomography

What is computed tomography?

Computed tomography, or CT, is a scanning technology that uses X-rays to create images of the inside of an object; it is widely used in hospitals for diagnosing internal injuries and diseases. CT is a powerful technique, which is capable of producing images cheaply and quickly, without causing any damage to the original sample. This means that it has many applications outside of hospitals, including the study of fossils (i.e. palaeontology). Medical scanners use low-energy X-rays and short exposure times to reduce the amount of radiation a patient receives, but non-medical (industrial) CT is not restricted in this way (when applied to inanimate objects like fossils). Therefore, such systems can be designed to penetrate dense layers of rock and generate very high-resolution images, resolving details less than one-hundredth of a millimetre in size.


Schematic diagram showing principles of computed tomography


How does it work?

During a typical industrial CT scan, a beam of electromagnetic radiation (i.e. X-rays) is emitted by an X-ray source, passing through the object under study while it is rotated around 180° or 360°. X-ray images are captured by a detector panel at regular intervals during this rotation, and these can be used to create hundreds or thousands of cross-section images, or slices, with the aid of a computer algorithm. The slices map the extent to which X-rays are attenuated – i.e. caused to lose intensity – by the sample; as a result, they distinguish between parts with different chemical compositions, which attenuate X-rays to differing degrees. This allows, for example, a fossil to be distinguished from the rock surrounding it, as long as they are made up of different minerals.


Different steps in the virtual reconstruction of a fossil


CT scans can be reconstructed as three-dimensional virtual fossils with the aid of special computer software. This typically involves generating binary (black/white) images based on the shades of grey in the CT slices. Here, the goal is to have all the parts of the images that represent the fossil turned ‘on’ (pure white pixels), with everything else (e.g. rock) turned ‘off’ (black pixels). If the fossil and the rock have very different grey levels in the slices (reflecting differences in their chemical composition), this process can be fairly straightforward, but considerable human input is usually needed. After the fossil has been virtually distinguished from the rock, it can be helpful to colour in different parts of the fossil, so that these can be individually examined in the final 3-D model.