About Computed Tomography (CT)

From the University of Texas CT Lab website:

Scientists have always been interested in looking at the internal structures of objects. Before the twentieth century, the only way that the structures of the insides of a physical model could be mapped was by physically cutting up the model (destroying it in the process).

With the invention of the x-ray, a non-destructive means of studying such objects became available. Unfortunately, radiograph images are difficult to interpret in three dimensions, with objects layered one on top of the other.

During the 1970s, this problem was solved by the development of computed tomographic or "CT" scanning. In this technique, a specific plane of an object rotating on a turntable is illuminated by a narrow X-ray beam. After passing through the specimen, the X-rays are recorded by an array of detectors. Digitized information on the attenuation of the X-rays passing through the illuminated plane at various angles and at various times during rotation is then used to mathematically reconstruct a two-dimensional density map of the sample. By sequentially imaging many two dimensional "slices" of a specimen, its three dimensional structure is recorded. In more sophisticated facilities, this sequence of slices may then be mathematically assembled into a three-dimensional density map that can be digitally manipulated. For example, artificial cross-sections may be generated in any plane, and sequential sections can be linked by means of animation software to simulate the passage of the specimen through a fixed plane. By this means, CT scanning non-destructively provides tomographic information superior to that obtained by physical serial sectioning.

The scientific usefulness of CT scans is strongly influenced by two issues: image resolution and the ease with which data can be exported and analyzed. Resolution is a function of the thickness of the slice illuminated by the X-ray beam. Medical CT scanning facilities generally obtain slice widths of 700 microns or 1400 microns. Although morphologists have used these with some success in studying large specimens, their images of smaller objects are generally poorly resolved. Additionally, such facilities usually store digital information in proprietary, non-exportable formats. Consequently, data analysis must be performed either from hardcopy or at the scanning facility. Despite these limitations, paleontologists, comparative anatomists as well as geologists have obtained useful data from such facilities.

During this decade, the limitations of medical scanners have been surmounted by the development of industrial CT scanners capable of slice resolutions of 20 microns and less. This is over two orders of magnitude greater than that available though medical facilities. Furthermore, the digital information generated by these machines is stored in standard exportable file formats. Consequently, data obtained in this way may be manipulated and published by the researcher in its native medium outside the scanning facility.


Link to the NPS Museum Management Program Home Page