DECT of the thorax can be performed using either single-source CT systems equipped with a rapid kV switching X-ray tube or dual source CT systems. With dual source CT systems, DECT is performed analogous to a conventional contrast-enhanced chest CT except that the tube current is distributed over both tubes, which are operated at different photon energies. The recommended spectral combination differs between the various generations of CT systems with the lower energy spectrum ranging from 80 to 100 kVp and the higher energy spectrum ranging from 140 to 150 kVp. In the more recent generations of dual source CT systems, the higher energy spectrum can be hardened by means of a tin filter in order to decrease the overlap in photon energies between both spectra. If the goal of performing DECT is the assessment of lung nodules, masses or lymph nodes, the examination is typically acquired in a venous phase to ensure adequate contrast enhancement of these structures although an arterial phase may be added depending on the clinical indication. In the delayed phase, the enhancement of the pulmonary parenchyma is minimal. If an assessment of pulmonary perfusion is desired, an additional dual energy CT pulmonary angiography (DE-CTPA) needs to be acquired, as described in more detail below.

Pulmonary nodules are frequently detected in chest CT. This includes CT staging examinations in patients with extra-thoracic malignancies, incidentally detected nodules in patients without an underlying oncologic diagnosis or nodules detected at lung cancer screening. The latter two scenarios are particularly challenging, since the pre-test probability of malignancy is relatively low. Several clinical risk factors as well as nodule size and characteristics have been shown to predict the likelihood of malignancy [1, 2]. Nevertheless, for an individual patient, these clues are often not sufficient to classify a nodule as either clearly benign or malignant. In this context, the ability of DECT to visualize and quantify the iodine uptake of pulmonary nodules has been explored as a means to differentiate benign from malignant pulmonary nodules [3]. DECT examples of malignant and benign pulmonary masses are shown in Figs. 4.1, 4.2, 4.3, 4.4, and 4.5, respectively. Chae and colleagues [4] evaluated the clinical utility of DECT for the classification of 45 solitary pulmonary nodules, which were confirmed as benign or malignant on the basis of percutaneous needle aspiration histology. In their study, the diagnostic accuracy for malignancy by using CT numbers on iodine-enhanced image series with a cutoff of 20 HU was comparable to that achieved by using the degree of enhancement. More recently, Hou and colleagues [5] compared the dual energy CT characteristics of 35 patients with lung cancers to those of 25 patients with inflammatory masses using a dual phase (arterial and venous phase) CT acquisition protocol. They found significant differences in iodine uptake between both groups. The iodine content in the center of masses in the venous acquisition phase (normalized to the aorta) had the highest discriminatory power with lower central iodine content in lung cancers compared to inflammatory masses. Kawai and colleagues [6] evaluated the feasibility of measuring iodine uptake of ground glass lesions by DECT. This study provided preliminary data that iodine uptake may be useful for the characterization of ground glass opacities as they found an increased iodine-related attenuation in adenocarcinomas compared to pulmonary hemorrhage or inflammatory changes. The presence and pattern of calcifications of lung nodules is another important feature that can suggest a benign or malignant entity. Virtual noncontrast images can be reconstructed from DECT data and have been found useful in differentiating iodine enhancement from calcifications in lung nodules (Figs. 4.2 and 4.3) [7].

The assessment of patients with a confirmed diagnosis of lung cancer represents a very different scenario. Here, the potential utility of DECT could be to assess tumor vitality and response to therapy. In addition, visualization of iodine uptake has been explored as a tool for the visualization and characterization of mediastinal lymph nodes in the nodal staging of lung cancer (Fig. 4.1). Schmid-Bindert and colleagues [8] have demonstrated that the iodine uptake of non-small cell lung cancers measured by DECT correlates with metabolic activity on FDG-PET. The correlation between metabolic activity at PET and the iodine-related attenuation at DECT was lower for thoracic lymph nodes compared to the primary tumor [8]. In a pilot study of ten patients with non-small cell lung cancer treated with the anti-angiogenic agent Bevacizumab, Kim and colleagues used DECT-derived tumor enhancement to assess therapy response [9]. In their study, the spectral information from DECT was considered particularly useful to discriminate between tumor enhancement and intratumor hemorrhage. Baxa and colleagues [10] retrospectively analyzed the DECT characteristics of mediastinal lymph nodes in patients with non-small cell lung cancer who underwent dual phase (arterial and venous phase) DECT before and after chemotherapy. They noticed that enlarged lymph nodes showed a higher arterial enhancement fraction compared to nonenlarged lymph nodes. In lymph nodes shrinking under chemotherapy, a significant decrease in arterial enhancement fraction was observed.

In patients with lung cancer, particular those with a centrally located tumor compromising the hilar vessels or bronchi, preoperative DECT has been used to predict postoperative pulmonary perfusion and ventilation. Chae and colleagues [11] found that DECT can predict postoperative lung function more accurately than scintigraphy. In their study, DECT was superior to scintigraphy for the depiction of perfusion and ventilation defects and assessing collateral ventilation in patients with centrally located lung cancers.

System-specific recommendations for DE-CTPA acquisition parameters have been proposed in the literature [22]. Both test bolus or bolus tracking techniques can be used. A slightly longer delay (5–7 s) than for conventional CTPA is typically chosen to allow for the contrast material to reach the pulmonary capillary system. An iodine delivery rate of approximately 1.6 g I/s has been found to result in optimal image quality for both morphological CTPA reconstructions and dual energy perfusion maps [15]. In order to minimize streak artifacts originating from dense contrast material in the superior vena cava, some institutions prefer a triphasic injection protocol in which the contrast agent bolus is followed by 30 mL of a mixture of saline and contrast agent, followed by a 50 mL saline chaser [23]. Choosing a caudocranial direction of acquisition further helps minimize artifacts from dense contrast material in the superior vena cava. As with conventional CTPA, deep inspiration prior to breathhold should be avoided, since the inflow of nonopacified blood from the inferior vena cava during deep inspiration can lead to an interruption of the contrast bolus and inadequate opacification of the pulmonary circulation. From the acquired low- and high-energy datasets, anatomical image series similar to a conventional CTPA can be reconstructed by mixing the data from both energy levels. In addition, specific postprocessing software is used to segment the pulmonary parenchyma and then display pulmonary iodine content as a color-coded parametric map (Figs. 4.6 and 4.7). By means of a different postprocessing algorithm, DE-CTPA data can also be used to visualize the iodine content within the pulmonary vessels to increase the conspicuity of emboli [24].