Compared to other imaging modalities, the distinctive strength of conventional CT imaging is its ability to acquire volumetric, morphological information of the patient's anatomy in a fast, reliable, and easy way. With modern CT systems, the patient can be scanned from head to toe in just a few seconds with an isotropic spatial resolution better than half a millimeter. CT has become the workhorse in diagnostic radiology. However, conventional CT acquisitions only reveal the patient's morphology; they do not provide any information about the chemical composition of the examined structures. Tissues having different chemical composition but the same x-ray attenuation coefficient will appear with the same grayscale value (Hounsfield value or CT number) in a CT image. Calcified plaques and iodine in a contrast-filled vessel are chemically very different; yet, depending on their density, they may show similar x-ray attenuation and be difficult to distinguish in a CT image.

This differentiation issue is not specific to CT but is a general problem of x-ray imaging. Looking at the grayscale values alone would not unambiguously tell us what portion is bone, and what portion is the metal ring. Only because of our knowledge of the displayed anatomical information are we able to differentiate between the bones and the ring.

Adding additional information about the examined tissues by data acquisition with different x-ray spectra may help overcome this limitation. The simplest approach is the use of two different x-ray energies - we call this dual-energy data acquisition. We will see later that for many of the relevant tissues and materials in a human body, dual-energy CT is sufficient to determine material-specific information so that a differentiation is possible, despite identical x-ray attenuation.