Micro-Computed Tomography

History

Computed tomography (CT) is a non-destructive technique that provides three-dimensional images of the internal structure of an object. The basic idea of this imaging technique goes back to J. Radon, who proved in 1917 that an n-dimensional object can be reconstructed from its (n-1)-dimensional projections. However, the mathematical basis for the actual CT image reconstruction was presented in two papers by Cormack in 1964 and 1965, respectively. About 10 years later, Hounsfield submitted a patent, describing the first CT scanner, which was then built in 1975. The possibility of non-invasively imaging three-dimensional sections of a human body was of such importance that Cormack and Hounsfield were awarded with the Nobel Price for Medicine in 1979.

Because CT provides an excellent contrast of bone to soft tissue, new medical developments of this technique were often driven by the bone research field. Feldkamp et al. were first to build a micro-computed tomography (microCT) scanner for the evaluation of the three-dimensional micro-structure of trabecular bone. At this stage, microCT was an experimental technique available to a few research groups only. However, with the presentation of the first commercially available bone microCT scanner in 1994, this technique became quickly a standard in bone research. Nowadays, microCT scanners are available from several manufacturers, who provide a whole palette of different scanner types targeting many applications that range from in-vivo measurements down to the analysis of bone tissue on a micrometer scale.

With the ongoing development of CT systems, this technique became available on many different levels of resolution, while always using the exact same physical working principal. Thus, CT is an excellent technique for investigation in a hierarchical fashion ranging from whole bodies down to the sub-cellular level (Figure 1). With the development of microCT, complementary techniques as well as new image processing algorithms and analysis techniques have evolved. These new techniques opened the field of microCT to many new applications.

Figure 1: Hierarchical imaging using computed tomography (CT). The technique can be used on a large scale with different resolutions, while always using the same physical principals. In highresolution CT (HR-CT) domain, normal X-ray tubes can be used as a source, whereas for microCT special microfocus X-ray tubes are required. The lower range of microCT as well as the nanoCT domain is currently best assessed using synchrotron radiation (SR). The images show from left to right, human hand (courtesy of Thomas L. Müller, ETH Zürich), trabecular bone structure, microcallus, murine cortical bone surface of a femur with internal vasculature (courtesy of Philipp Schneider, ETH Zürich).

The MicroCT Technique

The basic physical principal of computed tomography is the interaction of ionizing radiation, such as X-ray with matter, where, in the energy range typically used for CT imaging, the so-called photo-effect builds the main interaction mechanism. The photo-effect attenuates the photons proportional to the third power of the order number of the elements and inverse proportional to the third power of the photon energy. Thus, the actual attenuation not only depends on the material but also on the energy spectrum of the X-ray source. As an X-ray beam penetrates an object, it is exponentially attenuated according to the material along its path. The energy-dependent material constant appearing in the exponent of this attenuation formula is called the linear attenuation coefficient. It expresses the amount of radiation that is attenuated on an infinitely small distance, in which the final attenuation reflects the sum of all these local linear attenuations along the X-ray beam. Therefore, an X-ray projection (or X-ray image) represents an image of the sum of all local attenuations along the X-ray beam.

To produce a three-dimensional CT image, a whole set of such two-dimensional projections need to be acquired. In microCT, these projections are usually taken in a setup in which the source and detector are at a fixed position and the object is rotated around its long axis (Figure 2). The source is mostly either a microfocus X-ray tube or an insertion device of a synchrotron radiation facility and the detector is normally based on a CCD camera with a phosphor layer to convert X-ray to visible light. Since CCD cameras have a limited number of pixels, the projections are recorded in discrete points with a so-called sampling distance (distance between neighboring pixels) and a maximal number of samples (which may correspond to the number of pixels on the CCD). It can be shown that the number of projections taken over 180 degrees should be about twice the number of samples per projection to avoid aliasing artifacts. The two-dimensional projections can then be used to reconstruct a three-dimensional image. In this sense, CT images can be seen as images that represent linear attenuation coefficients.

Figure 2:Main components and working principle of a microCT scanner. A micro-focus X-ray tube emits X-ray, which is collimated and filtered to narrow the energy spectrum. The X-ray passes then the object and is recorded by a two-dimensional CCD array. A full scan involves a set of projections under different rotations of the object.