Brittany Dolph Dinneen, Dr. John A. Malko, Renee A. Stein
Conservators routinely document the structure of objects for the purposes of condition assessment, technical study, and treatment decision-making. X-radiography has provided conservators with the ability to generate images of otherwise inaccessible features such as closed cavities, internal armatures or other structural features of solid objects, in a nondestructive way. Computed tomography (CT) scanning provides an additional level of information, emitting x-rays in multiple planes and acquiring data at multiple angles to produce a three-dimensional reconstructed image, whereas the X-radiograph results in an image in which the three-dimensional information is superimposed into a two-dimensional format. In both X-radiography and CT scans image contrast, that is, areas of differing brightness, are related to the differences in X-ray attenuation (radiodensity) of the regions through which the x-rays travel. In X-radiography the superimposed nature of the image makes quantitation of radiodensity difficult. In CT scanning regions of the scanned object can be quantified by their radiodensity; this quantitation uses units called Houndsfield units, or CT numbers, which describe the radiodensity of a given volume relative to air (-1000) and water (0).
In spite of the proven capabilities of CT scans for imaging, however, scan output may still not provide enough visual evidence of morphology to characterize a material. Furthermore, visual interpretation of radiodensity is typically only semi-quantitative at best – allowing for only qualitative comparisons - and dependent on the parameters of the scan. We suggest that if the relationship between x-ray attenuation and material class is significant, the identification of a material or class of materials would be informed by comparing the assigned radiodensity of a determined region of interest to a table of known ranges for various cultural heritage materials (i.e. ceramics, stone, soil, clay, cellulosic organics, keratinous organics, etc.). Similar tables developed for medical diagnostics include CT numbers for fat, blood, muscle, gray and white brain matter, and different types of bone.
Potential applications include the characterization of materials in hidden cavities which may be unethical to open, or impossible to access, without irreversibly disturbing the exterior matrix. For example, many African power objects, such as minkisi, contain materials in bundles or otherwise hidden spaces. Other possible applications may be the material characterization of amulets hidden in mummy bundles and sealed opaque vessels with contents intact.