In the early 1980s, the first dual-energy application for a clinical CT system was introduced, allowing the assessment of bone mineral density from dual-energy data. In this approach, data were acquired using a kV-switching technique. Unfortunately, this led to several limitations, such as compromised acquisition protocols and reduced image quality. In addition, increased dose values were necessary, since only the kV could be adapted accordingly, and adaptation of the mA is necessary for optimal dual-energy processing. As a consequence, the overall mA level was increased to fulfill the requirements for the low-kilovolt beam, an increase that led to a substantial dose increase. These limitations, in combination with the limited separation of the spectra, reduced this technique to the field of bone densitometry measurements and prevented its general routine use. In addition, due to the competition with DEXA, the only remaining dual-emission application (bone densitometry) was in the end not migrated to other CT systems.
With the introduction of the first DSCT system, most of the above-mentioned limitations for dual energy were overcome, and a broad routine use of dual energy became feasible. In addition to the 'classic' applications, new applications such as the automatic removal of bone, tissue classification (e.g., gout and kidney stone classification), and the calculation of virtual noncontrast images were also viable. The issue of dose was finally solved with the introduction of a second generation of dual-source systems, since the improved systems allowed routine acquisitions without additional dose penalties.
In this chapter we describe the different approaches for processing dual-energy data from a clinical perspective. In addition, practical approaches and solutions for both image- and raw-data-based techniques are described and clinical examples presented.