Imaging of COPD

Abstract

Although imaging of patient with COPD in not recommended in clinical practice, many researches and drug trial include imaging studies in evaluation of COPD. In particular, thin-section volumetric CT covering the whole lung with following quantitative assessment has been accepted one of the essential components of studies. This chapter encompasses the whole aspect of CT in COPD, including basic CT physics, radiation dose consideration, diagnosis and visual assessment of COPD, fundamentals of quantitative assessment of emphysema, airway wall thickening, and air trapping and its clinical values. Assessment of other components of COPD includes chest wall changes, vascular changes, osteoporosis, and so on. In addition, new emerging imaging methods such as MRI, dual-energy CT, PET, and optical imaging to evaluate not only morphologic but also functional aspects of lung in COPD will be introduced. Functional imaging includes direct visualization of ventilation, perfusion, chest wall motion, and inflammation. The promising results of early studies will be introduced, highlighting the potential of imaging approach for the future studies.

The original version of this chapter was revised. An erratum to this chapter can be found athttps://doi.org/10.1007/978-3-662-47178-4_24

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Computed Tomography (CT)

CT Basic Physics

Computed tomography (CT) of the thorax is a very quick and the most advanced imaging technique. CT scanner is designed to use a form of gantry which allows rotation of the X-ray tube and the detector around the patient during breath hold. With the development of CT technology, single breath-hold scanning of the thorax can be achieved, and image reconstruction of sagittal and coronal planes is also feasible with minimal loss of spatial resolution. During the decades, the fan beam of a CT scanner is broadened along the Z-axis and two-dimensional detectors are developed, acquisition of a number of slices per rotation is possible. Recently developed multi-slice and multi-detector CT allows cone beam CT and volumetric imaging. After CT scan, the X-ray attenuation, called tissue density, is expressed as a Hounsfield Units (HU). The scale of HU values range from −1000 HU (attenuation value of air) to 3000 HU, and 0 HU corresponds to the attenuation of water (Fig. 8.1). Generally, normal lung has an attenuation value around −850 HU on inspiratory CT because normal lung attenuation reflects the mixed attenuation of intrapulmonary air and lung parenchyma.

Fig. 8.1

Hounsfield unit on CT images. The Hounsfield unit (HU) is a quantitative value for describing attenuation on CT images. Zero HU and −1000 HU are defined as attenuation of distilled water and air at standard pressure and temperature. CT images can be appropriately displayed according to target organs adjusting window width and level. Window width describes the range of HU on CT image and window level is the median HU of window width. In the thorax, lung, mediastinum, and bone window setting are usually used

Radiation Dose of COPD CT

The radiation dose during CT scan is presented as the gray or mGy unit, which is proportional to the amount of energy that the irradiated body part is expected to absorb. The Sievert (Sv) unit is used in the report of the effective dose. Regarding chest CT scan for evaluation of chronic obstructive pulmonary disease (COPD), acceptable low dose screening and standard dose of CTs can be performed at effective dose of approximately 2 and 7 mSv, respectively [1].

CT Protocol

Optimization of CT protocol and quality control of image acquisition are critical for assessment of COPD [2, 3]. Ideally, single CT protocol using dedicated single CT scanner and software system based on exactly the same parameters of image acquisition and reconstruction should be used. Nevertheless, it is impossible in most cases, particularly for multicenter trials. There are several important issues to be considered for the optimization of CT protocol which include kilovoltage setting, tube currents, respiration level, image thickness, reconstruction kernel, and so on (Table 8.1). Generally speaking, the volumetric CT acquisition obtained at maximal inspiration with standardized breathing instruction is essential for accurate COPD assessment using CT. Thin-section CT reconstructions (even or less than 1 mm in thickness) are required for proper characterization and quantification of emphysema and airway change in COPD. It has been known that CT estimates of emphysema severity increase as section thickness decreases and that higher frequency (edge-enhancing, sharper) image reconstruction kernel results in higher CT measurements of emphysema than lower frequency resolution (smoothing) kernel [4,5,6]. Low radiation dose CT scan allows the visual evaluation of emphysema, however, which trade off relatively high image noise, resulting in overestimation of emphysema extent. Relatively higher dose CT is also necessary for the accurate measurement of airway dimensions. Recently, further reduction of CT dose can be possible with the use of novel technique of iterative reconstruction [7,8,9]. However, it is not recommended for the quantitative assessment of COPD CT because the influence of changing reconstruction method has not been fully studied. Dose modulation technique to reduce radiation dose is not recommended. Expiratory CT is gradually regarded as an important part of COPD evaluation for the presence and distribution of air trapping on visual and quantitative assessment [10]. Expiratory CT scan may be obtained at relatively low dose after inspiratory CT acquisition. However, different noise level between inspiratory and expiratory CT is potential problem for the quantitative analysis. Expiratory scan can be obtained at the end of a tidal expiration, which corresponds to functional residual capacity, or after full expiration, which is in residual volume status.

Table 8.1 General principle of CT protocol for COPD evaluation

Diagnosis of Morphologic Change in COPD

Emphysema: Definition of CT Finding, Subtype, Pathologic Correlation

Pulmonary emphysema is defined as an abnormal permanent enlargement of the airspaces distal to the terminal bronchioles, which results in lower CT attenuation values than those of normal healthy lung areas. Although pulmonary function test (PFT) is useful for clinical severity assessment of COPD, it does not represent the type and range of heterogeneous pathophysiologic abnormalities in COPD [11, 12]. In addition, it tends to be relatively insensitive to both early stages and small changes in COPD [13]. CT scan may be ideal for the detection and characterization of COPD in that it allows for an in vivo analysis of morphologic characteristics and distribution of emphysema. Visual assessment is also useful to subtype the emphysema into centrilobular, panlobular, and paraseptal types. Centrilobular emphysema is the most common morphologic subtype of pulmonary emphysema. Pathologically, it begins near the center of the secondary pulmonary lobule (the most fundamental structural component of the lung containing parenchyma, airways, lymphatics, and vasculature) in the region of the proximal respiratory bronchiole. Parenchymal destruction starts in the center of secondary pulmonary lobule results in the characteristic apposition of normal and emphysematous lung (area of low attenuated destruction surrounded by normal tissue). On CT, small round low attenuated holes are evenly distributed with ill-defined borders in early stage and these low attenuated areas become confluent and inseparable with paucity of pulmonary vascularity in late stage (Fig. 8.2). Panlobular emphysema is characterized by permanent destruction of the entire acinus distal to the respiratory bronchiole, and its pathogenesis relates to alpha-1 antitrypsin deficiency. Parenchymal destruction involving entire acinus, which is contrast to centrilobular subtype, affects the lower lobes more severely (Fig. 8.2). Paraseptal emphysema tends to be limited in extent and occurs most commonly along the subpleural portion of the upper lung, often coexisting with other types of emphysema and fibrosis. Small focal lucencies, up to 10 mm in size, can be seen on CT (Fig. 8.2). Bullae or blebs, termed interchangeably, are focal regions of emphysema with no discernible wall, usually more than 1 cm in diameter at subpleural location. In some cases, they can be very large and may result in pneumothorax in COPD patients (Fig. 8.3).

Fig. 8.2

Emphysema subtypes on CT images. (a) Normal lung parenchyma. (b) Centrilobular emphysema. Scattered small focal lucencies (parenchymal destruction) in upper lobes, measuring more or less than 1 cm in diameter. (c) Paraseptal emphysema. Small focal lucencies, up to 1 cm in diameter, are located in subpleural ar8ea adjacent to the pleura and septal lines. (d) Panlobular emphysema. Secondary pulmonary lobules area completely replaced with emphysema with showing uniform and relatively homogeneous lucencies across parts of the secondary pulmonary lobules

Fig. 8.3

Large bullae and pneumothorax in a COPD patients. A 71-year-old male with COPD complaint of sudden dyspnea. (a) CT image at 3 month ago showed large subpleural bullae with collapsed central lung. (b) Left pneumothorax was noted with collapsed lung and large bullae

Airway Change: Bronchus, Trachea

Bronchial wall thickening, bronchial luminal irregularity, and bronchiectasis are commonly seen in patients with COPD with mixed findings of emphysema. Bronchial wall thickening is regarded as a sign of chronic bronchitis, easily identified in heavy cigarette smokers. Pathophysiologically, irreversible and progressive histologic changes in airways show diffuse hyperplasia of mucous glands associated with hypersecretion and bronchial wall thickening. Traditionally, bronchial wall is regarded to be thick if the diameter ratio of inner to outer lumen of bronchus is greater than 0.8. Nevertheless, the diagnosis of bronchial wall thickening of COPD with naked eyes on CT is subjective and still limited (Fig. 8.4). Moreover, it is virtually impossible to differentiate bronchial wall thickening of COPD from acute bronchitis or asthma. Bronchiectasis can be accompanied in COPD patients and may represent severe airflow obstruction [14, 15]. Trachea and main bronchus abnormalities can be visually defined on CT in COPD patients. Tracheobronchomalacia, saber-sheath deformity of trachea, and outpouching of trachea and main bronchus can be seen in advanced COPD (Fig. 8.5). Tracheobronchomalacia is traditionally defined as a more than 50% collapse of the trachea and main bronchus at end-expiratory CT. Saber-sheath deformity is seen as coronal narrowing and sagittal widening of the intrathoracic tracheal diameter. Tracheal outpouching is defined as a focal herniation of mucosa through the tracheal wall.

Fig. 8.4

Airway changes in COPD. (a) Example of bronchial wall thickening. Typical and severe thickening of wall of entire segmental bronchi in both lower lobes is noted. The diameter ratio of inner to outer lumen is smaller than 0.8. (b) Example of bronchiectasis. Marked dilatation of bronchial lumen in both lower lungs is noted. The inner luminal diameter of bronchus is greater than the diameter of the accompanying pulmonary artery. There are also loss of normal tapering with bronchial wall thickening

Fig. 8.5

Tracheobronchomalacia. Tracheobronchomalacia is defined as abnormal collapse of airway lumen on expiration. This abnormal finding can be seen in COPD patients. CT images on inspiration (a) and expiration (b) demonstrate severe lumen narrowing of the trachea on expiration (b) and CT images on inspiration (c) and expiration (d) also depicted severe lumen narrowing of both bronchi (d). In addition, severe air trapping is seen in both lower lobes

Air Trapping

It has been understood that small airway airflow resistance is the major site of obstruction in patients with COPD and precede the onset of emphysematous destruction in both centrilobular and paraseptal emphysema phenotypes of COPD [16]. Therefore, early detection and diagnosis of small airway disease in COPD is fundamental for early diagnosis of COPD. The small airways are referred to as airway lumen less than 2 mm, which cannot be visualized directly using even recently developed CT scanners. Thus, the finding of air or gas trapping, which appears to decrease lung attenuation on expiratory CT, can be used as indirect sign of small airway dysfunction in COPD because this finding is thought to be caused by early collapse of small airway on expiration. Expiratory CT is increasingly regarded as an essential tool for the evaluation of air trapping as an obstructive pattern of small airway disease in COPD patients. In many cases, especially in emphysema dominant COPD patients, the detection of small airway dysfunction may be hampered by the presence of emphysema because the lung density of emphysema area on expiration CT can be low without small airway dysfunction. In such situation, side-by-side comparison of inspiratory and expiratory CT images is necessary to detect lack of normal increase of lung attenuation and decrease of lung volume on expiration (Fig. 8.6).

Fig. 8.6

Air trapping assessment using inspiration- and expiration CT images. Air trapping should be assessed by side-by-side comparison of inspiration (a, b) and expiration (c, d) CT images. Expiratory air trapping can be defined as areas showing lack of normal increase of lung attenuation and decrease of lung volume on expiration image (c, d). The areas of air trapping are marked in arrows

Others: Chest Wall, Diaphragm, Heart, Pulmonary Vessel, and Bone

Morphologic changes secondary to pulmonary hyperinflation in COPD include chest wall deformity, diaphragmatic change, changes in the heart, and pulmonary vasculatures. Barrel chest deformity is the well-known chest wall deformity of COPD and depression of diaphragm indicates the flattening of the domes of the diaphragm due to hyperinflation of the lung in COPD (Fig. 8.7). As the progression of COPD, the heart tends to be more vertically oriented due to hyperinflation of the lung. In later stage, right ventricular and atrial dilatation, dilatation of central pulmonary arteries, and acute tapering of distal pulmonary vessels can be seen as a finding of pulmonary hypertension. Osteoporosis is also one of the systemic effects associated with COPD attributed by inactivity, COPD-related systemic inflammation, the use of systemic corticosteroids, and vitamin D deficiency. Bone fracture related to osteoporosis, in turn, may also reduce pulmonary function or even cause COPD exacerbations [17].

Fig. 8.7

Morphologic change of the diaphragm in COPD patients. Compared with CT images in a normal individual (a, b), coronal CT images (c, d) in a 74-year-old male with COPD shows depressed and flattened shape of the diaphragm. This change is due to hyperinflation of the lung in COPD

Severity Assessment of COPD

Extent of Emphysema: Visual Assessment

Visual assessment of COPD is relatively simple, cheap, and independent from variation of CT machine or reconstruction algorithms. Furthermore, comprehensive visual emphysema assessment of CT in COPD allows assessment of the pattern, subtype, regional location, and degree of emphysema. It also has an advantage for detecting lots of accompanying pathologic changes in the parenchyma as well as in the small and large airways. Visual assessment allows for the detection of early emphysema in asymptomatic smokers even before the development of airflow limitation, which is essential for the diagnosis of COPD. It is also useful to follow the progression of emphysema over time. For the optimal evaluation of emphysema with CT images in COPD patients, thin-section, high-resolution CT images should be used at recommended window settings (usually with a window level of −700 and window width of 1000–1500). On visual assessment, emphysema is classified as centrilobular, panlobular, and paraspetal emphysema (Fig. 8.2). The extent of emphysema has been assessed by using visual scoring system [18]. Typically, the extent of emphysema in each lobe can be assessed by using a six-point scale system: 0, 1–5%, 6–25%, 26–50%, 51–75%, and greater than 75% [19]. However, main limitation of visual assessment has been the relatively low inter-reader agreement [20, 21]. The inter-reader agreement was moderate for the presence or absence of emphysema and for the presence of panlobular emphysema; fair for the presence of centrilobular and paraseptal subtypes [22]. In an effort to improve the inter-reader agreement, usage of standardized reference images has been attempted with promising results [19].

Extent of Emphysema: Computer-Based Quantification

For the objective and reproducible assessment of emphysema, computer-based quantification method, the so-called quantitative CT (QCT), has been introduced. As briefly explained in “CT Basic Physics” section, emphysema area on CT with relatively increased air fraction in inspiration lung is shown as area of decreased CT attenuation approaching air attenuation of −1000 HU, compared with normal lung CT attenuation around −850 HU [23]. Accordingly, if a certain threshold value is applied, the area of emphysema with decreased CT attenuation can be objectively divided from normal lung area. This method is called as “density mask” and the threshold of −910 HU was initially applied [24]. Recent studies using thin-section, multi-detector CT scanners showed that the highest correlation between QCT and histology is found when the threshold set at −960 or −970 HU [25]. However, the lower the thresholds, the more sensitive the image noise; therefore, the threshold of −950 HU is now most commonly used. When the correction for lung volume changes influenced by degree of inspiration is applied, this quantitative method for emphysema is near perfectly reproducible method. The term of percent of emphysema (emphysema index, EI or %LAA₋₉₅₀) stands for the relative area of lung less than −950 HU (Figs. 8.8 and 8.9). Another method is using the nth cutoff percentile in the attenuation distribution curve using the CT attenuation at a certain percentile along the frequency histogram of pulmonary parenchymal attenuation (density value in HU under which n% of frequency histogram) (Fig. 8.8). This value is called as “percentile index,” and it is reported that it has an advantage on longitudinal evaluation and less sensitive to lung volume changes [26,27,28]. The first percentile value is much optimal for correlation with histology; however, it is known to be sensitive to an artifact from image noise and truncation effect. Instead, the 15th percentile threshold is commonly used [28, 29]. The last method is to assess the mean of the whole lung attenuation (mean lung density, MLD). Regional heterogeneity of emphysema can be assessed quantitatively to assess the regional distribution of emphysema. Most available QCT methods can divide each lung into upper, mid, and lower zones of equal height or volume, and ratios between the extent of emphysema in upper and lower lung can be computed. Newer methods can also permit segmentation of lobes to compute lobar volumes and extent of emphysema objectively (Fig. 8.9).

Fig. 8.8

Quantitative CT measurements of emphysema regarding LLA%, percentile index, and mean lung density (MLD). The concept of QCT measurement is illustrated in Fig. (a). Frequency distribution curves are plotted according to the apparent X-ray attenuation values. The threshold of −400 HU is to define lung area. Threshold (LAA %) technique uses a predefined threshold of HU (i.e., –950 HU) is chosen and the percentage of lung less than this value can be calculated. Contrary to this, the percentile index method uses a certain percentile point (i.e., lowest 15th percentile) and the HU value for that percentile is calculated. The last index is the mean of lung density (MLD). Left side shift of frequency curve in patient with emphysema is demonstrated in Fig. (b). As a result of the shift, LAA % increases and percentile index decreases in patients with emphysema. The area of emphysema can be overlaid on the CT images to highlight the emphysema lesion (c, d)

Fig. 8.9

Lobar-specific quantification of emphysema. Example of software providing automatic segmentation of lobes and quantitative assessment of extent of emphysema at the whole lung and lobar level

Comparison Between Visual Assessment and CT Quantification

In the assessment of emphysema in COPD patients, although QCT measures correlate with the severity of visual CT assessment, the level of correlation is only moderate [22]. Especially in less severe categories of emphysema, visual assessment by radiologists tends to be usually underestimated for the extent of emphysema when compared to QCT measures, while in those with more severe emphysema, the radiologists tend to relatively overestimate the emphysema extent [30]. Therefore, visual assessment and QCT measures should be used complementarily and performed independently for the assessments of severity of emphysema.

Correlation Between the Extent of Emphysema and Clinical Parameters

Visual, subjective assessment of the emphysema using contiguous 10-mm thick CT started in 1986, there were significant correlations between CT visual scores and macroscopic emphysema. However, even with the development of high-resolution CT, visual grading assessment is not really a quantitative measure but just grading the degree of emphysema according to categories of emphysema severity. Recently, using pulmonary lobe-by-lobe visual assessment, severity of emphysema correlates quite well with physiologic parameters (FEV₁ and FEV₁/FVC) and GOLD stage [19]. The correlation coefficient ranges between 0.67 (for GOLD stage) and −0.74 (for FEV₁/FVC) and notably the range of correlation coefficients are similar to the correlations between extent of emphysema on QCT and each physiologic parameter (0.62 for GOLD stage and −0.70 for FEV₁/FVC). However, inter-reader agreement regarding severity of emphysema on visual assessment tends to be variable, so QCT is preferred for assessing disease severity of emphysema. Moreover, QCT measurements have shown to correlate better than visual CT assessment with macroscopic measurement of emphysema [21].

Regional Heterogeneity of Emphysema

Severity and distribution of emphysema differs in lung regions such as core (inner) vs. rind (outer), upper vs. lower, even among each lobe (Fig. 8.10). There have been several reports regarding regional variation of emphysema. Basal distribution of emphysema is associated with greater impairment of FEV₁ but less impairment of gas exchange (PaO₂) and alveolar-arterial oxygen gradient than the apical distribution of emphysema [31]. Emphysema areas on CT are more often found in the inner segment of the lung than in the outer segment and the extent of emphysema in inner segment of the lower lung in QCT is much more clearly correlated with airflow limitation than those in outer segment [32]. In another report applying the slope of the EI in the upper-lower, anterior-posterior, and central-peripheral direction in both side lung, the heterogeneity of emphysema distribution in anterior-posterior and upper-lower direction was independent determinants of FEV₁ and FEV₁/FVC and the lower and posterior regional dominant emphysema is associated with a decrease in FEV₁ and FEV₁/FVC [33]. Regional assessment using QCT helps in selecting candidates for lung volume reduction surgery (LVRS) and provides rationale for the mechanisms of improvement after LVRS (Fig. 8.11). The extent of emphysema of the upper-rind region of the lungs is a significant predictor for improvement of pulmonary function after LVRS [34].

Fig. 8.10

Regional heterogeneity. CT images with color mapping (a, b) show different predominance of emphysema in two patients with similar emphysema index (17.56 and 23.27%). This regional heterogeneity can also be quantified by emphysema on each lobes (c)

Fig. 8.11

Effect of lung volume reduction procedure. (a) CT image in a 61-year-old male showed severe and regional heterogeneity of emphysema. Endobronchial L was performed to collapse the RUL which was most severely affected by emphysema. (b) After the procedure, RUL was near totally collapsed on CT and pulmonary function of this patient much improved (FEV1, 0.46 L —> 0.82; mMRC, Gr4 —> 1)

Airway Change: Visual Assessment

Although visual assessment of large airways is a subjective process, the presence of airway wall thickening or dilation of large airways can represent bronchial inflammation with remodeling, and it also contributes to the symptomatic exacerbation in COPD patients. Bronchial wall thickening is commonly found in heavy cigarette smokers with a sign of chronic bronchitis (Fig. 8.4). Bronchiectasis is also a common finding in COPD associated with severe airflow obstruction and risk for COPD exacerbation [14, 15, 35].

Airway Change: Computer-Based Quantification

With the current development of available software permitting multiplanar reconstruction of airways from thin-section volumetric datasets, CT scan seems to be well positioned to become the method of choice to noninvasively measuring airway wall dimensions of luminal diameter and wall thickness to the level of segmental and subsegmental airways. The simple analysis of “full-width at half maximum” algorithm is commonly used to evaluate the airways including absolute measures (bronchial luminal diameter or area, bronchial wall thickness or area, and total bronchial area) and relative measures (bronchial wall area %, WA%: 100 × (wall area)/(lumen + wall area)) (Fig. 8.12). Variable software algorithms to define the boundaries of airway wall have been proposed and tested. Another commonly used measure is the square root of wall area of a hypothetical bronchus having internal perimeter of 10 mm (Pi10) [36].

Fig. 8.12

Quantification of airway wall thickening using computerized method principle for full-width-at-half-maximum (FWHM) method. In the attenuation profile along an outward flowing ray from the luminal center-point through the airway wall, the inner and outer airway wall boundaries are assumed halfway to the maximum on the lumen side and halfway to the minimum on the parenchymal side, respectively. (a) After the detection of boundaries, the airway wall area can be highlighted with color overlay (b). This method is called as full-width-at-half-maximum method and is most common method to quantify wall thickness. The airway dimensions of the whole airway trees can be assessed using automatic airway segmentation, centerline extraction, followed by airway quantification (c)

Correlation Between Airway Measures and Clinical Parameters

Many studies showed that patients with the greatest WA% had the lowest FEV₁ expressed as a percent predicted [37]. The WA% has been considered as the most commonly employed metric for clinical investigation, and there are modest correlations between airway WA% and lung physiologic impairment [10, 38]. Moreover, central airway remodeling apparent on CT may reflect the distal histopathologic remodeling of the small airways, so the greater the central airway wall thickening, the more small the airway disease [39]. Moderate correlations (−0.56 < r < −0.62) between airway wall measurements and airflow obstruction (FEV₁ and FEV₁% predicted) have been reported and stronger correlations were noted when only small airways were analyzed [40]. In recent report, bronchial wall thickening as well as severity of emphysema measured on QCT is associated with exacerbation frequency, independently; bronchial predominant and emphysema predominant subtypes of COPD can be defined [41].

Other Large Airway Changes in COPD

Among the large airway changes in COPD, Saber-sheath trachea is not an uncommon finding in COPD (Fig. 8.5). It defines as the ratio of the sagittal to coronal diameter is greater than 2 and the extra-thoracic portion of the trachea is not narrowed. By comparison with normal healthy persons, COPD patients show that this tracheal morphologic change of the elongation of the sagittal diameter correlated with the severity of emphysema and QCT indices, reflecting airflow limitation and air trapping [42]. Furthermore, expiratory tracheal collapse in obese COPD patients shows greater quality of life impairment and worse exercise performance than expected based on functional measures [43].

Small Airway Disease: Visual Assessment

The small airways are referred to as airway lumen less than 2 mm, those cannot be visualized using current CT scanners. The presence of air trapping on expiratory CT scan can be identified as an indirect sign to evaluate small airway remodeling. However, accurate discrimination between emphysematous area and air trapping area is difficult and challenging, and even normal healthy person can show minimal air tapping area in both basal lobes. Recently, there have been introduced several methods to evaluate the degree of pathologic air trapping in COPD patients using QCT.

Small Airway Disease: Computer-Based Quantification

With the usage of expiratory CT images, quantitative assessment of air trapping is possible. However, there are severe inborn limitations when only expiratory CT is used because it is impossible to separate trapped air area from air remaining in emphysematous spaces. Moreover, air trapping phenomenon can also be seen in healthy smokers and healthy individuals with normal lung physiology. However, even with these drawbacks, many studies have evaluated the presence of air trapping in COPD by assessing the area fraction of the lung lesion having CT values lower than the threshold of −856 or −850 HU (LAA_exp856 or LAA_exp850) in expiration [44]. With this simple method, they reported that high correlations were noted between LAA_exp850 and FEV₁/FVC and FEV₁ percent predicted. As an effort to overcome this single threshold method of combining air trapping and emphysema quantification into single measure, quantifying air trapping outside the emphysematous area is possible through the density-based quantification method [45, 46]. With excluding emphysema portion of all voxels with attenuation lower than −950 HU from inspiration and expiration scans and calculating the relative volumes for whole lung with attenuation value less than −860 HU on each inspiratory CT (inspiratory relative volume_{< −860 HU}) and expiratory CT (expiratory relative volume_{< −860 HU}), the relative volume change between −860 and −950 HU can be calculated as follows: Relative volume change _{< −860 HU} (%) = expiratory relative volume _{< −860 HU}—inspiratory relative volume_{< −860 HU}. Results from this method show that air trapping correlates with lung physiologic parameters significantly (r = 0.50–0.80). Other methods for measuring air trapping have been addressed as an index of air trapping including the ratio of inspiratory to expiratory lung volume (E/I-ratio_LV) and the expiratory to inspiratory ratio of mean lung density (E/I-ratio_MLD) [47, 48]. E/I-ratio_MLD correlates with clinical parameters of COPD such as BODE-index (0.48 < r < 0.68) and E/I-ratio_LV shows almost perfect correlation with E/I-ratio_MLD (r = 0.95, p < 0.001). All of these values have a limitation that it can’t represent regional distribution of air trapping. Recently, new approach of air trapping assessment has been proposed [49, 50]. By using software techniques, anatomical correspondence of lung region between inspiration and expiration CT images can be assessed and both images can be co-registered. By direct assessment of density difference between inspiration and expiration CT, the density change map can be generated and the areas with decreased density change can be defined as the area of air trapping. By using this method, in addition to the global assessment, regional assessment of air trapping is possible (Fig. 8.13) [49, 50].

Fig. 8.13

Subtraction image from co-registered inspiratory and expiratory images. Image maps (c, d) derived from co-registered inspiratory (a) and expiratory (b) images depict changes in lung attenuation from inspiration to expiration. Using image registration technique, the expiration CT image is deformed to match with the corresponding anatomical area on inspiration CT (c). By comparing CT attenuation between inspiration CT and registered expiration CT, area of air trapping with little change in CT attenuation can be extracted and visualized in color overlay (c). This process is applied in the whole lung and coronal distribution of air trapping can be visualized (d)

Other Components of COPD: Correlation with Clinical Parameters

Texture-Based Emphysema Assessment

With an effort to discriminate various obstructive lung diseases from normal lung, more sophisticated automatic classification system based on the texture and shape features of CT images has been introduced. Using this method, further differentiation of lung areas, for example, into normal lung, small airway disease, centrilobular emphysema and panlobular emphysema, is possible. This quantification method showed comparable correlation with the pulmonary function test results when compared with conventional density-based quantification [51, 52]. This method can also be used for the assessment of combined pulmonary fibrosis.

Pulmonary Vascular Change

Pulmonary vascular change is also one of the characteristic features of COPD and the extent of emphysema, rather than airway obstruction, is responsible for pulmonary endothelial dysfunction in COPD [53]. After volumetric CT scans of the lung, pulmonary vasculature was automatically segmented from the parenchyma using software [54, 55] (Fig. 8.14). With the usage of QCT measuring the cross sectional area (CSA) of small pulmonary vessels (sub-subsegmental level, CSA less than 5 mm²), the total CSA of small pulmonary vessels in COPD shows strong (negative) correlation with the extent of emphysema (%LAA₋₉₅₀), whereas weak correlation with airflow obstruction [56]. The anatomic extent of emphysema instead of airway obstruction is responsible for impairment of pulmonary vascular structure. Moreover, the percentage of the total CSA for the lung area is significantly higher in airway dominant phenotype than emphysema dominant phenotype.

Fig. 8.14

Vascular subbranches of lung. (a) Extraction vessels in CT and (b) Vascular tree reconstruction; random coloring; each color represents one vascular branch and mediastinal region is cropped

Osteoporosis

Osteoporosis should be considered as one of the important pathology of COPD, because it may cause vertebral compression fractures, which can also deteriorate FEV₁ and decrease in vital capacity. Actually, the loss of vertebral bone mineral density on CT is closely related to the severity of emphysema showing many risk factors of low BMI, decreased activity, systemic inflammation, and use of corticosteroids [17, 57, 58] (Fig. 8.15). There has been reported that the decrease in thoracic vertebral bone mineral density is greater in patients with a history of exacerbations than in those without a history of exacerbations. Indeed, osteoporosis progression should be checked in COPD patients, especially in those with a history of frequent exacerbations [59].

Fig. 8.15

Thoracic bone density on CT image. Vertebral bone mineral density on CT is closely related to the severity of emphysema. (a) CT image of a 59-year-old male with 6.2% of emphysema index shows 206.7 HU on thoracic vertebra. (b) CT image of a 79-year-old male with 52.2% of emphysema index shows 93.0 HU on thoracic vertebra

Chest Wall and Diaphragm

COPD is also characterized by progressive impairment of respiratory function and dysfunction of respiratory muscle. There have been reports that the depletion of peripheral muscle mass is a better predictor of mortality than BMI in patients with COPD [60, 61]. Thoracic respiratory muscles are unique and crucial for alveolar ventilation and weakness of respiratory muscle results in dyspnea and respiratory failure associated with mortality in COPD patients. It is reported that intercostal mass and intercostal attenuation measured by QCT are significantly correlated with FEV₁ and extent of emphysema of QCT measurement [62]. A decrease in thoracic muscle mass with increasing intercostal fat is associated with worsening of COPD (Fig. 8.16). Hyperinflation in COPD makes diaphragm to be flatter and shorter. As the progression of COPD, breathing becomes gradually more dependent on the thoracic intercostal muscles than diaphragm (Fig. 8.7).

Fig. 8.16

Thoracic muscle mass on CT image. Thoracic muscle mass and intercostal fat are associated with severity of COPD. (a, b) CT images of a 75-year-old male with 7.4% of emphysema index shows more prominent thoracic muscle mass than CT images of a 62-year-old male with 54.8% of emphysema index

Atherosclerosis

In COPD patients, reduced FEV₁ has known to be an increased risk factor for cardiovascular diseases and mortality, independent of smoking [63, 64]. In other words, systemic inflammation in COPD patients may accelerate the rates of cardiovascular disease, and this degree or status of atherosclerosis may be associated with impaired lung function and emphysema in COPD patients. There has been an attempt to demonstrate the association of the total amount of calcification in coronary artery, thoracic aorta, mitral and aortic annuli, and the extent of emphysema on QCT and lung physiology [65]. Calcium score as a measurement for degree of atherosclerosis shows weak but significant correlation with volume fraction of emphysema on QCT, FEV₁/FVC, and diffusion capacity, independent of age, BMI, and smoking amount. The degree of atherosclerosis is associated with impaired lung function and the extent of emphysema.

Imaging Studies of Treatment Monitoring and Disease Progression

Predicting Tool in Treatment Outcome

COPD is a heterogeneous condition featuring both parenchymal destruction (emphysema) and small airway disease (obstructive bronchiolitis); its relative contributions vary in each COPD patient. Previous studies already have suggested that different spirometric response patterns to bronchodilator exist in patients with obstructive lung disease showing improvement in expiratory flow (FEV₁) or lung volume (FVC) [66, 67]. Therefore, QCT measurements can be used as a longitudinal evaluation of treatment monitoring based on the fact of a significant correlation between QCT measurement indices and lung physiologic indices. Notably, in the assessment of different spirometric response patterns to bronchodilator treatment, the extent of emphysema in QCT measurement shows a significant negative correlation with postbronchodilator FEV1 change and the E/I-ratio_MLD also shows a significant positive correlation with postbronchodilator FVC change [68]. In case of lung volume reduction therapy in regions with severe emphysema, QCT can be used as a predictor of improvement of lung function after surgical or bronchoscopic approaches [34].

Disease Progression

There have been several efforts to evaluate the progression of emphysema using QCT measurement and QCT is useful in demonstrating the change in extent of emphysema directly (Fig. 8.17). However, the main drawback of QCT measurement is a variation of inspiration level on each CT scan. Studies showed that the change in the 15th percentile CT density after the correction of lung volume difference was found to be more sensitive as an index of progression compared with measures of physiology or healthy status [69]. Recently, new single unified approach using a voxel-wise imaging analysis can be used in diagnosing disease extent and phenotype of COPD, detailed spatial distribution and location. This method allows us to distinguish the relative contributions of functional small airway disease component and emphysema in COPD in the course of disease progression [49].

Fig. 8.17

Follow-up emphysema quantification (LAA, %) in same subject. Initial emphysema index (EI, %) was measured to 41.8% on CT image (a) and matched color coding image (b). Three years hence, EI has increased and measured to 56.2% on CT image (c) and matched color coding image (d)

Quality Control and Standardization

Emphysema Quantification

There are several sources of variation in quantification of emphysema in COPD including scanner, software, and patient factors. Thinner slice thickness and the lower CT dose setting result in overestimation of emphysema extent on QCT due to increasing image noise [4]. Inter- and intra-scanner variation due to calibration error and beam hardening effects should also be considered, and there have been attempt to demonstrate that density correction of volumetric CT data based on air (reference value: −1000 HU in tracheal air or outside patient air) will improve the correlation between emphysema quantification and pulmonary function test [70, 71]. This correction method may be useful to decrease the variation of measured results when multiscanners are involved. Second, a smooth reconstruction algorithm is usually recommended for the emphysema quantification using QCT because a strong or overenhancing algorithm can result in overestimation of emphysema [6]. Regarding patient factors, variation in lung volume, which is influenced by degree of inspiration, can be a major source of variation in clinical practice. Measurements of emphysema can be different at varying inspiration levels. However, quantitative measurement of differences in emphysema would not be significant when scans are obtained above 90% of vital capacity [72]. Current smoking status or coexisting air trapping or parenchymal fibrosis can also alter the quantitative emphysema measurements. Measured extent of emphysema in current smokers appears to be lower than those in former smokers, probably due to increased attenuation induced by smoking-related infiltration of inflammatory cells in the lungs in current smokers [73, 74]. Therefore, for accurate and precise quantitative assessment of emphysema extent, it is important to consider and control all the factors discussed above.

Airway Quantification

In quantitative airway measurements, considering factors about the source of variation are similar to emphysema measurements. Because airway is small, it is easily influenced by partial volume averaging and reconstruction algorithm than in quantitative measurement of emphysema. Appropriate radiation dose to overcome image noise during reconstruction and submillimeter section will reduce the variation in airway measurements. In addition, there are varieties of suggested software algorithms to measure airway dimension, resulting in different measurement values. Accordingly, usage of same software method is essential in multicenter trial or for following up the patients.

Magnetic Resonance Imaging (MRI)

The MRI can obtain morphologic and functional imaging without ionizing radiation. However, there are some reasons of difficulty in imaging the lung with MRI. The lung consists of the low density tissue, so it contains a relatively small number of protons which generate signal in MRI. In the lung, these are fast decay of signal due to susceptibility artifacts from countless air-tissue interfaces. The fast imaging, triggering, and gating techniques are needed due to respiratory, vascular, and cardiac motions. The major advantage of MRI is combination of morphological and functional lung imaging, such as perfusion, ventilation, blood flow, gas exchange, and respiratory motion, with high spatial and temporal resolution [75].

Morphologic Evaluation of COPD

Simply speaking, there are two different types in COPD. The airway type relates to chronic bronchitis and airflow obstruction. The emphysema type shows the parenchymal destruction with severe airflow obstruction and distal airspace enlargement. The morphologic evaluation in COPD using MRI is always compared to CT. MRI is technically more challenging due to hyperinflation and loss of lung tissue [76]. So, there are only a few publications on morphologic evaluation of COPD using MRI.

Imaging Technique

Many MRI sequences can be used for visualization of the lung: balanced steady-state free precession (bSSFP), volumetric interpolated breath-hold examination (VIBE), half-Fourier-acquired single-shot turbo spin-echo (HASTE), and ultra-short echo time (UTE) [77,78,79]. Three-dimensional (3D) T1-weighted gradient-echo sequences are used for the assessment of mediastinum and lung parenchymal tissues. T2-weighted fast spin-echo with half-Fourier acquisition sequence can visualize bronchial wall thickening and mucus plugging. Respiratory, vascular, and cardiac motions can be overcome by using fast imaging, gating, and triggering techniques. Half-Fourier acquisition or ultra-short echo times are recommended.

Emphysema

It is a major role of T1- and T2-weighted images to differentiate inflammation from muscular hypertrophy, edema, and mucus plugging in bronchial wall [80]. Emphysematous change of lung cannot be easily diagnosed by a loss of signal. However, hyperinflation can be easily detected by increased lung volume and reduced blood volume. There is one study about the change of signal intensity of lung parenchyma between inspiration and expiration MRIs, which is correlated with FEV₁ [81]. The MR signal can be improved and emphysema can be quantified by using the UTE pulse sequences [82]. Using fast radiofrequency (RF) excitation pulses, compressed sensing and parallel imaging in UTE pulse sequence, MR signal decay, and motion artifacts can be minimized [83]. UTE pulse sequences improve contrast-to-noise ratio, signal-to-noise ratio, and signal intensity with strong relationship between signal intensity and tissue density. Using UTE pulse sequence, pulmonary emphysema [84, 85], lobar fissures and airways [86], inflammation and peribronchial abnormalities [87] can be estimated.

Airway

There are several factors, such as bronchial level, diameter, wall thickness, and signal from bronchus, to detect bronchiectasis [88]. MRI usually visualizes central and peripheral bronchiectasis and central bronchi, whereas poorly visualizes normal peripheral bronchi. Using 3D volume interpolated gradient-echo sequence (VIBE) with high spatial resolution, the airway can be visualized [89]. T2-weighted sequences can visualize inflammation, mucus, edema, and fluid collections. Active inflammation can be represented by increased fluid, which shows high signal of the bronchial wall on T2-weighted sequences. Inflammatory activity has relation with contrast enhancement of thickened bronchial wall on contrast-enhanced T1-weighted sequence. Therefore, airway inflammation can be visualized by high signal on T2-weighted and contrast-enhanced T1-weighted sequences [90]. In contrast, mucus plugging on peripheral bronchi shows high signal intensity of fluid content on T2-weighted sequence without contrast enhancement on T1-weighted sequence on MRI.

Perfusion MRI

Using perfusion MRI, perfusion information of lung can be acquired without ionizing radiation. One of the advantages using perfusion MRI in COPD is combination of perfusion and morphologic information about parenchymal destruction and cause of perfusion changes. Several imaging techniques have been introduced.

Imaging Techniques

For assessment of pulmonary vasculature and perfusion, both of non-contrast-enhanced and contrast-enhanced sequences are available. Non-contrast-enhanced perfusion MRI can be acquired using arterial spin labeling technique, which is to mark a specific part of spins magnetically using radiofrequency (RF) excitation [91]. Using the electrocardiogram (ECG) gating technique, signal differences between systolic phase and diastolic phase can make perfusion images of the lung without contrast injection [92]. However, one of the limitations of ECG-gated perfusion MRI is that the image subtraction process is sensitive to misregistration due to bulk respiratory motion [91]. Contrast-enhanced 2D and 3D MRI, which is based on dynamic acquisition of lung tissue during contrast injection, can assess lung perfusion and quantify pulmonary perfusion [93, 94]. The advantage of contrast-enhanced perfusion MRI is high signal-to-noise ratio [95]. For evaluation of whole lung perfusion during peak enhancement period, 3D technique should be needed. For improvement of spatial resolution and reduce of the total acquisition time, k-space sampling techniques such as parallel imaging techniques or echo sharing techniques can be used [96, 97].

Quantification

Using MR perfusion technique, pulmonary blood flow can be assessed quantitatively [98, 99]. The indicator-dilution theory using the maximum of signal intensity and the temporal course of the signal change is base of quantification of pulmonary perfusion. If linear relation can be assumed between the concentration of contrast agent and the signal, concentration-time curves can be made by conversion of signal-time curves. The relationship among the perfusion parameters—pulmonary blood flow (PBF, mL/100 mL lung tissue/min), pulmonary blood volume (PBV, mL/100 mL lung tissue), and mean transit time (MTT, s0) is as follows:

$MTT=\frac{PBV}{PBF}$

Normalizing the area under the tissue concentration-time curve to the integral of the arterial input function can calculate PBV. Figure 8.18 shows quantification images of lung perfusion using perfusion MRI technique.

Fig. 8.18

Lung perfusion quantification image. Using perfusion MRI technique, PBF map (a), a PBV map (b), and an MTT map (c) are generated

COPD Studies Using Perfusion MRI

In COPD, chronic inflammation is thought to lead to intimal wall thickening and smooth muscle hypertrophy of pulmonary arteries. Hyperinflation and air trapping can make hypoxic vasoconstriction. Perfusion MRI shows high accuracy in detecting perfusion abnormalities in patients with emphysema [100, 101]. The perfusion abnormalities in COPD usually show a low degree of inhomogeneous contrast enhancement, especially in the area of severe emphysema [102] and decreased peak signal intensity. In COPD patients with severe emphysema, visual assessment of perfusion using 3D perfusion MRI shows high agreement with parenchymal destruction [103]. With quantitative analysis, decreased perfusion parameters on MRI correlate with worsening of FEV₁/FVC and increased emphysema index on CT [104]. Quantitative perfusion MRI in COPD shows decreased value and heterogeneous change in mean pulmonary blood flow (PBF), pulmonary blood volume (PBV), and mean transit time (MTT) than those in normal volunteer [105].

Ventilation MRI

For assessment of lung ventilation, several methods of MRI, such as oxygen-enhanced MRI and hyperpolarized noble gas MRI, are developed. Repeated or time-resolved measurements of lung dynamics can be obtained using ventilation MRI.

Oxygen-Enhanced MRI

Oxygen-enhanced MRI can visualize lung ventilation. Oxygen couples to hemoglobin and is present as dissolved oxygen in blood during oxygen exchange between capillary beds and alveoli [106]. Paramagnetic property of deoxyhemoglobin makes little T1 shortening effect with T2* shortening effect [107, 108]. Dissolved oxygen makes shortening of T1 relaxation time of blood in pulmonary vein due to paramagnetic property of oxygen [107]. This shortening leads to increased signal intensity on oxygen-enhanced MRI [109, 110]. Several sequences, such as centrically reordered phase-encoding scheme on HASTE sequence and single-shot rapid acquisition with relaxation enhancement (RARE) or HASTE sequences, can make oxygen-enhanced MRI [111]. Respiratory gating techniques are preferred because pulmonary physiology and physiopathology can be affected by breath hold despite decreased misregistration. Oxygen-enhanced MRI is used to assess ventilation abnormalities in pulmonary emphysema [112]. Regional changes in ventilation on oxygen-enhanced MRI reflect the regional lung function [113]. Figure 8.19 shows ventilation of patients with severe COPD using dynamic oxygen-enhanced MRI. FEV₁ and DLco well correlate with slope of oxygen-enhancement time-course curve and degree of oxygen-enhancement, respectively. Dynamic oxygen-enhanced MRI reflects DLco and provides diffusing capacity maps [112].

Fig. 8.19

Oxygen-enhanced MRI in COPD patients. (a, b) A 75-year-old female smoker with mild COPD (a: thin-section coronal multiplanar reformatted (MPR) CT image shows centrilobular emphysema in the left upper lung field, b: relative enhancement (RER) map from oxygen O₂-enhanced MRI demonstrates heterogeneously decreased oxygen-enhancement in the both lung, especially left upper lung filed.) (c, d) a 56-year-old female smoker with severe COPD (c: thin-section coronal MPR CT image demonstrates panlobular emphysema in bilateral upper and middle lung field. d: RER map from oxygen O₂-enhanced MRI demonstrates heterogeneously decreased oxygen-enhancement in the both lung, especially upper lung fields.) Image courtesy of Yoshiharu Ohno, Kobe University Graduate School of Medicine

Hyperpolarized Noble Gas MRI

Using hyperpolarized noble gas MRI with ³He or ¹²⁹Xe gas, ventilation MRI can be acquired [110, 114, 115]. These techniques visualize the ³He or ¹²⁹Xe gas in airway and airspaces, so it can be used for regional mapping of airflow and assessment of diffusion in airspace [116, 117]. For these techniques, specialized laser equipment and specialized RF transmitter and receiver coils are mandatory. Four different techniques are used for hyperpolarized ³He MRI [118]. Static ventilation imaging generally uses 2D or 3D fast low-angle single-shot or bSSFP sequences [119]. Diffusion imaging uses gradient-echo pulse sequence with bipolar diffusion-sensitizing gradient waveform between the excitation RF pulse and data acquisition [120]. Dynamic ventilation imaging uses the ultrafast pulse sequences such as interleaved spiral pulse sequence with good balance between spatial and temporal resolution [121]. Oxygen partial pressure imaging uses the paramagnetic effect of oxygen on polarization of ³He [122]. Single-acquisition and single breath-hold technique improves temporal resolution and reduces error due to second breath hold. MRI with this technique can directly measure the regional ventilation and perfusion distribution [123]. Because ¹²⁹Xe is naturally richer than ³He and ¹²⁹Xe MRI shows comparable quality to ³He MRI with recent advances in polarization and imaging methods; many methods of ³He are translated for ¹²⁹Xe. However, ¹²⁹Xe MRI differs from ³He MRI due to difference of gas distribution. In stage III COPD, ventilation defect volume (VDV) was sensitive to minimal changes during short-term follow-up [114]. Percentage of ventilation volume was significantly different among three groups (healthy volunteers, healthy asymptomatic smokers, and COPD patients). COPD patients are separated from healthy subjects by apparent diffusion coefficient (ADC) map [124]. Using diffusion-weighted hyperpolarized ¹²⁹Xe MRI, ADC was significantly correlated with PFTs (FEV₁, FEV₁/FVC, and DLco) [115]. Ventilation defect percentages in ³He MRI show significant correlations with ventilation defect percentages in ¹²⁹Xe MRI and ventilation defect percentages show strong correlations with FEV₁ [125]. ADC can measure airspace size sensitively. In COPD, airspace dimensions are increased compared to non-smokers [126]. ADCs in emphysema show regional variations and significantly larger than those of healthy volunteers, which is homogeneous [127].

Fourier Decomposition MRI for Combined V-Q Imaging

Non-contrast-enhanced ventilation and perfusion MRI, known as Fourier decomposition MRI, uses a short echo dynamic SSFP acquisition with subsequent compensation for respiratory motion by using non-rigid image registration [128, 129]. Peaks at the respiratory and cardiac frequencies can be identified by spectral analysis of the image time series. Deformation of lung parenchyma and pulmonary blood flow leads to regional proton density change, which is related to amplitude of these peaks [130]. With image post-processing, ventilation- and perfusion-weighted maps are generated for regional assessment of lung function from a single acquisition series [131].

Dynamic Respiration MRI

Diaphragmatic geometry is affected by hyperinflation of the lung. Dynamic respiration MRI with fast-acquisition technique can visualize complex interaction between chest wall and diaphragmatic motion with high spatial and temporal resolution [132]. Using dynamic respiration MRI, the change in lung volume during the respiratory cycle can be assessed. In emphysema patients, motion of the diaphragm and chest wall is reduced, irregular or asynchronous [133]. Emphysema in lower lung shows significant correlation with diaphragmatic flattening, abnormal chest wall motion, and severe airflow limitation [134]. Although both normal and paradoxical diaphragmatic motion is restricted by severe hyperinflation, the paradoxical diaphragmatic motion shows significant correlation with hyperinflation [135].

Other Imaging for COPD

Dual-Energy CT

Introduction of Dual-Energy CT

Dual-energy CT refers to CT that uses two different energy spectra (usually 80 and 140 kVp). With the knowledge about the X-ray attenuation change of a particular substance at two different X-ray energies, the material differentiation and elemental decomposition of tissues are possible in dual-energy CT [136, 137]. Therefore, with single contrast CT scanning at two different energies, the CT images used for structural evaluation are created by combination of the CT images from the low- and high-energy CT data, and the material-specific images such as iodine map or xenon map for functional evaluation are created by the material-decomposition algorithms in the postproccessing of dual-energy datasets (Fig. 8.20). Three different designs of dual-energy CT for the acquisition of dual-energy data have been proposed. Firstly, dual-source CT system has two separate X-ray tubes and two corresponding detectors, which are placed with an angular off-set of 90° on the rotating gantry. Each X-ray tube can be operated at different kilovoltage and milliamperage settings. Secondly, in rapid kVp switching, the single X-ray source is used and X-ray tube electronically switches the tube voltage between higher energy and lower energy in about 0.5 ms. Lastly, the energy-sensitive sandwich detector is now commercially available. This system uses a layered detector and single X-ray tube with the polychromatic spectrum. The layered detector comprises two layers: a thin top scintillator that absorbs low energy photons and a bottom scintillator that absorbs the higher mean energy photons.

Fig. 8.20

Iodine perfusion imaging using dual-source CT. CT image data is generated in dual-energy acquisition mode of dual-source CT. Axial CT image is obtained at 140 kVp and 50 mAs from A-tube and at 80 kVp and 210 mAs from B-tube. And conventional CT image (approximate 120 kVp image) is generated from combination of the 140 and 80 kVp datasets. Iodine image is obtained with extraction of iodine component with material-decomposition theory in post-processing. Fusion image with conventional CT image and iodine image is generated for the evaluation of lung parenchymal perfusion

For COPD patients, which is characterized by airway obstruction and emphysematous alveolar destruction, pulmonary vascular changes are also produced, which are characterized by hypoxic vasoconstriction, numeric reduction, and endothelial dysfunction of the small pulmonary arteries [138,139,140]. These characteristic anatomic changes influence and impair alveolar gas exchange, and the uneven distribution of alveolar ventilation and pulmonary blood flow (V/Q mismatch) is the most important cause of arterial hypoxemia in the COPD patients [141, 142]. Therefore, evaluating COPD patients should focus on not only the extent and severity of anatomic destruction of the lung parenchyma but also the functional changes and impairment such as alveolar ventilation changes or parenchymal perfusion changes.

Perfusion Dual-Energy CT

Dual-energy CT-derived iodine map represents the iodine content of the capillary bed, i.e., pulmonary blood volume at the time of CT scanning rather than pulmonary blood flow. However, it has been demonstrated that the similarity of the dynamic CT-derived pulmonary blood flow and dual-energy CT-derived pulmonary blood volume [143]. Therefore, pulmonary blood volume, assessed with dual-energy CT, can be used for the evaluation of lung perfusion as a surrogate for dynamic CT-derived pulmonary blood flow with simpler protocol while maintaining quantitative similarity (Fig. 8.21).

Fig. 8.21

Iodine perfusion image with dual-energy CT in COPD patient. (a–c) Dual-energy CT in COPD patient with mild emphysema. (d–f) Dual-energy CT in COPD patient with severe emphysema. From iodine perfusion CT using dual-energy CT, (a, d) weighted conventional CT image and (b, c, e, f) iodine perfusion map are generated. Lung parenchymal destruction in COPD patient can be assessed on high-resolution conventional CT image (a, d), and anatomically matched parenchymal perfusion information can be evaluated on perfusion map (b, c, e, f). In the area of confluent emphysema of both lower lobes, parenchymal perfusion is decreased (displayed with blue color)

In COPD, the regional lung perfusion impairment occurs due to the hypoxic vasoconstriction of the area with decreased ventilation and reduction of the pulmonary capillary by the chronic inflammation of pulmonary artery [144]. Moreover, alveolar surface destruction is accompanied by the reduction of the pulmonary capillary bed. Thus, considerable attention has been paid to the evaluation of lung perfusion alterations in COPD patients because the severity of parenchymal destruction and the alteration of lung perfusion determine the functional effect of emphysematous changes. Lung parenchymal perfusion has been assessed by perfusion scintigraphy, single-photon emission CT and MR. However, the distribution of perfusion impairment does not match with the area of parenchymal destruction [145]. Dynamic multi-detector CT perfusion imaging also can provide the regional perfusion. It has been demonstrated that smokers with subtle CT findings of centrilobular emphysema and normal findings at spirometry has increased regional heterogeneity of lung perfusion compared with never smoked subjects and smokers with normal CT image [146]. However, dynamic multi-detector CT perfusion imaging necessitates a central high-pressure bolus of contrast material and scanning a limited axial extent of the lung during a cardiac-gated scan.

Pansini et al. have demonstrated that regional alteration of lung perfusion can be assessed by dual-energy CT, matching parenchymal destruction in 47 smokers with predominant emphysema [147]. Moreover, Lee et al. have shown that the contrast-enhanced dual-energy CT can be used for the quantification of emphysema and regional perfusion evaluation by using the virtual non-contrast images and iodine map, simultaneously [148]. Assessing the distribution of pulmonary emphysema and anatomically matched parenchymal perfusion information is particularly applicable in the patient and target lobe selection for lung volume reduction surgery or bronchoscopic lung volume reduction. Data from National Emphysema Treatment Trial with more than 1000 patients undergoing lung volume reduction surgery showed that lung volume reduction surgery reduces mortality in patients with upper-lobe predominant emphysema only if there is low perfusion to the upper lobe on scintigraphy [149]. Park et al. used dual-energy CT with lung perfusion imaging for target lobe selection of bronchoscopic lung volume reduction by endobronchial valves. In that study, the target lobe was selected, if it was most hyperinflated and least perfused, and if it had no collateral ventilation with other lobes on perfusion image with dual-energy CT [150].

Ventilation Dual-Energy CT

In clinical practice, the evaluation of COPD severity is based on the result of pulmonary function test; however, pulmonary function test provides global status of lung function and does not show the regional distribution of functional abnormality. For evaluation of regional lung parenchymal ventilation, radionuclide scintigraphy or MR is used, but it is limited by its low spatial resolution. CT also can be used to depict lung ventilation with inhalation of xenon gas [151, 152]. Xenon is a radio-opaque gas and xenon gas concentration in alveolar space can be measured based on the attenuation changes on CT image (Fig. 8.22). However, because of variability in baseline lung attenuation between images due to misregistration artifacts and different respiration levels, the accurate measurement of lung ventilation function is limited, triggering great attention in the simultaneous structural and functional evaluation with single CT scanning accessible to dual-energy CT. Two stable gases, xenon and krypton, with high atomic numbers (54 for xenon and 36 for krypton) are eligible for ventilation imaging with dual-energy CT. For xenon ventilation imaging with dual-energy CT, the patient usually inhales 30% stable xenon (a mixture of 30% xenon and 70% oxygen) for 1 min to 1 min 30 s with use of a xenon gas inhalation system (Zetron V; Anzai Medical, Tokyo, Japan).

Fig. 8.22

Xenon ventilation image with dual-energy CT in COPD patient. (a–c) Dual-energy CT in COPD patient with mild emphysema. (d–f) Dual-energy CT in COPD patient with severe emphysema. (a, d) Conventional CT image and (b, c, e, f) xenon ventilation map image obtained from xenon ventilation dual-energy CT. On weighted average image (a, d), centrilobular emphysema and bronchial wall thickening in both lower lobes are noted. (b, c, e, f) In right upper lobe and left upper lobe, decreased xenon ventilation is identified, while xenon ventilation is preserved in left upper lobe posterior segment

The first clinical report with xenon ventilation imaging with dual-energy CT was reported by Chae et al., investigating eight healthy volunteers and four patients with COPD. The authors have demonstrated the direct visualization of the degree of xenon gas enhancement in the lung parenchyma as a color overlay on a conventional thin-section chest CT image by material decomposition [153]. Park et al. performed two phase (wash-in and wash-out phase) xenon ventilation dual-energy CT in 32 patients with COPD. And regional quantified value of xenon enhancement in abnormally low attenuating lung area on xenon-enhanced images with wash-out phase showed inverse correlation with pulmonary function test, and it showed better correlation with pulmonary function test than CT attenuation parameters [154]. Similar approaches were also reported in asthma patients. Investigating 22 asthma patients, Chae et al. have demonstrated the ventilation defects that appeared on xenon ventilation imaging with dual-energy CT in asthma patients with severe airflow limitation and airway wall thickening. And the extent of the ventilation defects on xenon-enhanced CT showed correlations with parameters of pulmonary function test [155]. Demonstration of the reversibility of airflow obstruction after inhalation of bronchodilator has been reported, and it suggested that xenon-enhanced dual-energy CT may be feasible for visualizing the changes of airflow in response to drugs in asthma patients [156,157,158].

Stable krypton can be an alternative to xenon for ventilation imaging with dual-energy CT due to its high atomic number, and lack of toxicity and anesthetic properties. Hachulla et al. have firstly reported the krypton ventilation imaging using dual-energy CT in COPD patients. Single CT acquisition covering the whole thorax was performed after inhalation of a mixture of 80% krypton and 20% oxygen with five respiratory maneuvers through the mask. The maximum level of krypton enhancement within the lung was 18.5 HU, which is lower than that reported with xenon, with an average maximum degree of xenon enhancement of 23.78 HU, is sufficient to detect ventilation abnormalities [159]. This single static approach in phantoms and volunteers also has been reported recently using xenon gas after a single vital capacity inhalation [160]. The single static evaluation delivers a lower radiation dose and a few or single inhalation method is more easily implementable for radiologists and patients, without side effects. However, different ventilation dynamics can be evaluated regarding the scanning method and gas inhalation method.

Combined Ventilation and Perfusion Assessment with Dual-Energy CT

Pulmonary parenchymal perfusion change or ventilation impairment was evaluated with dual-energy CT separately in COPD patients. However, the pulmonary parenchymal ventilation and perfusion are changed concurrently, and the imbalance between ventilation and perfusion is the important characteristics in the patients with COPD. Investigating ten patients with various diseases from an anesthesiological intensive care unit, Thieme et al. have reported the potential of dual-energy CT to provide both pulmonary ventilation and perfusion imaging [161]. Zhang et al. applied combined ventilation and perfusion imaging with dual-energy CT in patients with suspected pulmonary embolism [162]. Combined ventilation and perfusion imaging with dual-energy CT in patients with COPD has not yet been reported (Fig. 8.23).

Fig. 8.23

Combined ventilation and perfusion assessment with dual-energy CT. From combined xenon ventilation and iodine perfusion CT using dual-energy CT, the conventional virtual non-contrast CT image (a), xenon ventilation/iodine perfusion map (b), ventilation map (c) and perfusion map (d) are generated. Lung parenchymal ventilation, perfusion and ventilation-perfusion imbalance with high-resolution anatomic CT information can be simultaneously evaluated with combined ventilation and perfusion dual-energy CT

Nuclear Medicine Imaging

Scintigraphy, Single-Photon Emission Computed Tomography (SPECT)

Both planar scintigraphy and SPECT image the distribution of radiotracer which is introduced into the body, and the emitted radiation from radiotracer is detected by external detectors. In contrast to scintigraphy which forms a single two-dimensional image, analogous to a planar X-ray scan, SPECT provides three-dimensional imaging about the distribution of a radiotracer by combining scintigraphic and computed tomographic technique, and allows the functional information from SPECT to be easily combined with the high-resolution anatomic information from CT. Perfusion scanning is generally performed using 99 m-technetium-labeled macroaggregated albumin (99 m Tc-MAA), which lodges in the pulmonary circulation after peripheral injection. In an animal study using pigs, perfusion SPECT has been shown to be more sensitive than HRCT to detect mild physiologic changes of elastase-induced pulmonary emphysema [163]. Moreover, Suga et al. have been demonstrated that perfusion abnormalities on breath-hold SPECT-CT fusion image can better reflect the lung pathophysiology than the emphysema index on morphologic CT scan [164]. And there have been several intervention studies in COPD patients using perfusion scintigraphy and SPECT to predict and measure clinical success. As mentioned earlier, assessing the lung parenchymal perfusion has been particularly applicable in target lobe selection for lung volume reduction surgery or bronchoscopic lung volume reduction. Perfusion scintigraphy has been used in National Emphysema Treatment Trial with more than 1000 patients undergoing lung volume reduction surgery for selection of target lobe [149]. And perfusion scintigraphy was also useful for selection of target lobe for endobronchial valve therapy in advanced emphysema patients, and patients having heterogeneous emphysema with a low baseline target lobe perfusion benefited from endobronchial valve therapy [165]. Although scintigraphy and SPECT have constitutional problems, low spatial resolution and long image acquisition time, to date, they are widely applied due to their availability in many centers. Ventilation scintigraphy and SPECT use two types of inhalation radiotracers: gaseous radioisotopes or radiolabeled particulate aerosols. 81 m Kr and 133 Xe as gaseous radioisotopes have been used for ventilation scintigraphy and SPECT, and several studies with gaseous radioisotopes have demonstrated ventilation heterogeneity in COPD [166, 167]. And for radiolabeled particulate aerosol, Technegas (99 m Tc-labeled, aerosolized ultrafine carbon particle, approximately 200 nm diameter) is usually used in patients with COPD due to its small particle size [168], even in the presence of severe airflow obstruction [169]. With Technegas, the inhomogeneities in ventilation on scintigraphy and SPECT in COPD patients have been visualized and quantified [170, 171].

Ventilation/perfusion SPECT can also be applied for the evaluation of the imbalance between ventilation and perfusion in the patients with COPD. Jogi et al. have reported significant correlation between total reduction in lung function assessed with ventilation/perfusion SPECT and spirometric lung function and emphysema severity on CT in patients with COPD [172] (Fig. 8.24). Suga et al. have described that quantitative analysis of V/Q distribution by SPECT and the standard deviation and kurtosis of the V/Q profile could be adequate indicator for the severity of lung V-Q imbalance causing gas exchange impairment in patients with emphysema [173].

Fig. 8.24

Ventilation/perfusion SPECT. Patient with centrilobular emphysema on HRCT. Ventilation/perfusion SPECT shows uneven ventilation and perfusion. Extensive areas with reverse as well as matched ventilation/perfusion defects are found. Better perfused areas correspond to areas which are well ventilated (black arrows) and to better preserved parenchyma on HRCT (white arrows). Other areas appearing as hot spots on ventilation images are poorly perfused (dotted arrows). (Reprinted with permission, from reference 172)

Positron Emission Tomography (PET)

Regional ventilation and perfusion also can be evaluated with PET using isotope ¹³N₂ gas dissolved in saline solution. Brudin et al. have reported that high V/Q tended to be more common in subjects with an emphysema dominant subtype, whereas low V/Q was more common in those with a bronchial inflammation dominant subtype. Spatial heterogeneity of lung perfusion also has been described with ¹³N₂ saline PET, and the regional heterogeneity in perfusion has been increased in patients with mild COPD compared to healthy controls, after adjusting for regional changes in lung tissue density and ventilation. [174]. These results suggest that regional perfusion changes may precede lung parenchymal destruction in COPD. Therefore, this imaging method may serve as an early biomarker for COPD.

Recent studies have been focused on systemic inflammation as a consequence of COPD, and patients with COPD are found to have higher levels of systemic biomarkers of inflammation and worse cardiovascular risk profile and prognosis [175,176,177]. Fluorine-18-fluorodeoxyglucose (¹⁸F-FDG) is the most commonly used PET radioisotope, and it depicts increased glucose metabolism. In COPD patients, ¹⁸F-FDG has been used to demonstrate both pulmonary and systemic inflammation. The pulmonary inflammation as ¹⁸F-FDG uptake in COPD patients was significantly higher than in asthmatic and normal subjects [178]. Pulmonary ¹⁸F-FDG uptake in usual (PiM phenotype) COPD patients was greater compared to patients with alpha-1 antitrypsin deficiency-associated emphysema (PiZ phenotype) and healthy controls, and it was correlated with FEV₁ [179]. Coulson et al. have evaluated the aortic inflammation as a measure of systemic inflammation using ¹⁸F-FDG PET in seven COPD patients, five metabolic syndrome patients, and seven ex-smokers. Aortic inflammation in COPD patients was intermediate between ex-smokers and metabolic syndrome patients [180].

Optical Coherence Tomography (OCT)

OCT is an imaging method utilizing the refraction of light waves as it passes through tissues, and measures the lung structure with endobronchial approach. A fiberoptic probe with a near-infrared optical probe is introduced into the airways via a bronchoscope and the reflected light by the tissue is detected and reconstructed into an image. The spatial resolution of OCT is much higher than CT with the ability to resolve structures up to micrometers and distinguish between different tissue types within the airways. Thus, the main strength of OCT in COPD is the ability assessing small airway morphology in vivo. Coxson et al. have demonstrated an excellent correlation between airway lumen and wall area measured by OCT and by CT [181]. Furthermore, OCT airway dimensions measured at the fifth generation bronchi showed a stronger negative correlation with the subject’s FEV₁ than CT, and had greater sensitivity in detecting changes in wall measurement than CT measurements. And OCT also provided data on airway wall morphology and subepithelial remodeling and collagen deposition. While OCT is not widely applicable to date, the novel ability on directly assessing small airway disease in COPD patients may allow increasing utilization of OCT in research and clinical practice.