ABSTRACT
Dual Energy Computed Tomography (DECT), a special case of general-purpose multispectral CT, has emerged as a valuable adjunct to single-energy CT in multiple clinical applications in neuroradiology. Although multiple implementations of DECT technology exist, the physical principles underpinning its clinical deployment require acquisition of low- and high-energy images for the same anatomy. Technologies used for DECT include: (1) acquiring two sequential scans at different x-ray energies (sequential DECT); (2) splitting the x-ray beam in a high- and low-energy halves (twin-beam DECT); (3) using two x-ray source/detector pairs simultaneously (duel-source DECT); (4) rapidly alternating beam energy during scanning (kVp-switching DECT); (5) using energy sensitive detectors (sandwich-detector based DECT); and (6) sorting detected photons into multiple energy bins (photon-counting multispectral CT). This chapter will briefly discuss the various architectures used for implementing DECT.
Low- and high-energy images from DECT acquisitions are post-processed using models that incorporate the dependence of x-ray attenuation on photon energy as dictated by the photoelectric effect and Compton scattering. Assuming that each voxel consists of only two types of preselected materials (e.g., water and iodine), DECT data can be processed to do material decomposition. For example, one can generate virtual non-contrast and iodine-only images enabling differentiate between contrast staining and intracranial hemorrhage. Similarly, one can automatically remove bone from CT angiograms using the difference between energy dependence of bone and iodinated-contrast. The material discrimination methodology is quantitative and may be used to quantify materials. For example, one can measure the amount of iodine present in tumor. The dependence on photoelectric effect and Compton scattering may also be used to generate virtual monochromatic images at any preselected beam energy level. This capability may be used to reduce metal and beam-hardening artifacts, to enhance soft tissue contrast of brain parenchyma, or to enhance visualization or contrast material at low concentrations. This chapter will describe post-processing techniques used in DECT and illustrate them with useful and promising applications in neuroradiology.
