Apparatus and Computer X-ray Tomography: Visualization of Intrinsic Structure, Evaluation of Performance and Limitations

In this paper we compare confocal apparatus tomography and computer tomography for medical use. Two aspects are illustrated: the process of image visualization and the dose deposition. The latter aspect is a damage factor of the object under investigation; it is why the data collection time is evaluated. Two principal tomography schemes are described and the expressions connecting the collected signal value with the object's internal structure are given. Based on the expression, the process of image visualization is described and the dose deposition is calculated for both cases. The object's classification is introduced to choose the optimal experimental scheme for the object under study. INTRODUCTION The use of X-ray microdiagnostic systems in biology and medicine has grown, covering such areas of investigations as the tomography of small animals (Asadchikov et al. 2004), the tomography of labeled cells (Schneider et al. 2001), the three-dimensional evaluation of biocompatible materials (Muller et al. 2001), the microtomography for the analysis of osteointegration around implants (Bernhardt et al. 2004) etc. Fluorescence microtomography combines fluorescence analysis with X-ray tomographic techniques and enables multielemental observation (Golosio et al. 2003). The terms "apparatus" and "computer" describe two modes of the fluorescence collection during the experiment. In the first case a parallel (Takeda et al. 1995) or a confocal (Zaitsev et al. 2002) collimator is placed in front of the energy-dispersive detector window to decrease the detector solid angle and to localize the fluorescence generation zone from where the signal is collected. The registered signal is a function of the detector focal spot position. In the second case there is no space localization. The experimental set-up repeats the X-ray computer tomography parallel scheme but the fluorescence is registered in addition to the absorption signal. The registered signal is now a function of the X-ray microbeam position and the rotation angle (Simionovici et al. 1999). To compare the advantages and limitations of apparatus tomography and computer tomography we have analysed the image visualization process. The data manipulation procedures and the signal collection time (radiation dose) are evaluated for both cases. CONFOCAL APPARATUS TOMOGRAPHY WITH X-RAY MICROBEAM The scheme for confocal apparatus X-ray fluorescence tomography is presented in Fig.1. An X-ray microfocus beam (Snigireva et al. 2003) illuminates a small volume of the sample. That part of the sample generates fluorescence which is collected by an energy-dispersive detector through the confocal collimator (Fig. 2) placed in front of the detector. The collimator (Fig.2) manufactured by Microelectronics Technology (Zaitsev et al. 2004) features a microfocus spot size. The intersection of the microbeam and the collimator focus localizes a zone from where the fluorescence is collected. The sample mounted on the sample-stage is translated in X-Y-Z directions with an applicable offset. Lateral resolution of the technique is determinated by the scanning step, the X-ray beam size and the collimator focal spot size, appropriately. Usually the scanning step is equal to the collimator focus size and a pixel has the same size (Fig.1). Below we consider pixels rather than voxels to simplify the considerations. For each "focal pixel" (ifo pixel on Fig.1) a fluorescence spectrum is registered. Each spectrum is mathematically processed to obtain a discrete set of values. Fig.1. Scheme of the confocal apparatus X-ray Fluorescence Tomography. Collimator Object ifo pixel