1. Introduction

Chronic obstructive pulmonary disease (COPD) is the 4th leading cause of death in the world, and one of the leading causes of morbidity resulting in substantial economic burden on healthcare worldwide [1].

COPD is characterized by persistent airflow limitation caused by a combination of small airway disease (obstructive bronchiolitis) and pulmonary parenchymal damage (emphysema). Although COPD can develop in non-smokers, cigarette and other types of smoking is the most common recognized risk factor in the development of COPD. Exposure to various organic and inorganic dusts, chemical agents and fumes, such as from coal and wood burning, together with genetic predisposition, lung development abnormalities, accelerated aging, bronchial hyper-reactivity and low socio-economical status are additional risk factors for development of COPD [1]. Cumulative exposure to inhaled noxious substances and other risk factors over decades is believed to induce a modified chronic inflammatory response and altered repair mechanisms in the lung, resulting in a cycle of repeated injury and repair in the airways, lung parenchyma and vasculature. Over time this may lead to progressive airflow limitation, air trapping and parenchymal destruction, despite cessation of causal agents, such as smoking [1].

Essential components of a clinical diagnosis of COPD are spirometry, presence of symptoms and exposure to risk factors. COPD usually presents as chronic and progressive dyspnea, cough and mucous production. Spirometry testing is required to confirm COPD diagnosis. A post bronchodilator ratio of the forced expiratory volume in 1 s (FEV1) to forced vital capacity (FVC) below 70% (FEV1/FVC < 0.70) is diagnostic of persistent airflow limitation. Post bronchodilator FEV1 is an indicator of the severity of airflow limitation and is associated with an increased risk of acute exacerbations and death [1].

Diagnosis of COPD based on clinical information, including spirometry data, does not distinguish among different subtypes of COPD, which comprise emphysema, chronic bronchitis, asthma and bronchiectasis. Emphysema is defined as “an abnormal, permanent enlargement of the airspaces distal to the terminal bronchioles, accompanied by destruction of the alveolar walls, without obvious fibrosis” [2], whereas the remaining COPD subtypes result in small airway disease. From an imaging perspective, these subtypes can however often be distinguished, although it is important to emphasize that different COPD entities may coexist. Furthermore, imaging often allows characterization of emphysema types, yielding a more precise diagnosis.

This review focuses on the appearance of various subtypes of emphysema on chest radiographs (CXR) and computed tomography (CT). Radiographic features of the pulmonary diseases are generally described in terms of the degree of attenuation of the x-rays, with high attenuation being white and low attenuation being black in conventional radiology images. “Opacity” and “shadowing” refer to areas that are less dark than normal lung parenchyma should appear. “Lucency” appears as darker areas.

2. Emphysema classification

Morphologically emphysema is classified into three major subtypes which can coexist: centrilobular, paraseptal and panacinar. Such classification takes into account the affected portion of the acinus – the terminal respiratory unit of the lung located distal to the terminal bronchiole. Additionally, emphysema in which the enlarged airspaces are over 1 cm in diameter is referred to as “bullous emphysema”. Each of these subtypes is associated with underlying causes, as shown in Table 1. Distinguishing these subtypes is easier by CT than chest radiography, but becomes more challenging in advanced stages of the disease, even by CT.

Underlying cause Type of Emphysema	Tobacco Smoking	Connective tissue disorders	Alpha 1-Antitrypsin deficiency	IV talc injection	Heroin, cocaine and cannabis smoking	Hypo-complementemic Urticarial Vasculitis syndrome
Centrilobular	√
Paraseptal	√	√			√
Panacinar			√	√		√

Table 1.

Different types of emphysema versus underlying cause.

Centrilobular or centriacinar emphysema is the most common type of emphysema and the one most associated with cigarette smoking. Pathologically, the central portion of the acinus (i.e. the proximal respiratory bronchioles and their associated alveoli) is destroyed [3]. On CT, centrilobular emphysema may appear to arise from the center of the secondary pulmonary lobule [4]. It usually affects the upper lobes and superior segments of the lower lobes, with the central lung being more affected than the periphery. The reasons for this distribution are not fully understood and may include differential perfusion, immunological make-up, dust clearance mechanisms, pleural pressure and lymphatic flow of the affected regions.

Mild-to-moderate centrilobular emphysema is characterized by multiple round or oval low-attenuation areas, usually several millimeters in size, predominantly scattered in the upper and inner lung zones (Figure 1). These small areas of low attenuation have no definable walls (as opposed to lung cysts) and are bordered by normal lung parenchyma; however they may coalesce and form larger lucencies, with well-defined and thin walls formed by compressed adjacent lung parenchyma separating them from normal lung parenchyma [5].

Figure 1.
Forty nine year-old male with prior history of smoking (35 pack-year) and exposure to distillates and nickel in pharmaceutical lab with centrilobular and paraseptal emphysema. (a) PA CXR shows increased lucency bilaterally predominantly in the upper lung zones. Note the absence of vessels in the outer third of the lung in the bilateral upper lobes. Incidental note of a linear scar in the right mid lung. (b) Axial CT image allows better appreciation of the combined centrilobular and paraseptal emphysema. The areas of centrilobular emphysema appear as punched out lesions in the central portion of the lungs (yellow arrows). Notice the typical subpleural and upper lobe predominance of paraseptal emphysema (black arrows). Paraseptal emphysema is so severe that the subpleural interlobular septa are either not visible on this 1 mm thickness CT or partially destroyed. The patient subsequently underwent bilateral lung transplantation.

Although severe centrilobular emphysema may be difficult to distinguish from other subtypes, its distribution is helpful in differentiating it from panacinar emphysema (Figure 2). Also the presence of preserved lung parenchyma around large airways and vessels, anatomically located in the perilobular portion of the secondary pulmonary lobule, is a clue that emphysema has originated in the centrilobular portion [4].

Figure 2.
Seventy nine year-old male with very severe COPD (FEV1 0.85 l, 29% predicted; FVC 2.9 l, 74% predicted; FEV1/FVC 29%. Axial (a) and coronal (b) CT images demonstrate severe centrilobular emphysema mimicking panlobular emphysema. The upper lung predominance is the best clue allowing to differentiate it from panlobular emphysema. Also note a lesser degree of hyperinflation compared to panlobular emphysema such as in cases of alpha-1 antitrypsin deficiency (refer to Figure 5). The destruction of the secondary pulmonary lobules is so severe that the attenuation of the lung in the left upper lobe is equal to that of the column of air in the trachea (−980 HU); this was confirmed by impaired gas exchange (DLCO 4.6 ml/min/mmHg, 20% predicted).

Distal acinar or paraseptal emphysema typically affects the subpleural upper lungs including perifissural and peribronchovascular lung parenchyma. This type of emphysema may not be associated with smoking and is usually not associated with airflow obstruction. Pathologically, the distal portion of the acinus – alveolar ducts and sacs, is destroyed. On CT, it is characterized by multiple subpleural and peribronchovascular areas of low attenuation, ranging from a few millimeters to 1 cm in diameter (with occasional bullae formation), separated by thin intact interlobular septa (Figure 3). It should be distinguished from honeycombing, which presents with much thicker walls, is multilayered, more frequent at the lung bases and is associated with fibrosis and architectural distortion [3]. Blebs and bullae can result in pneumothorax in young patients [3, 4] (Figure 4).

Figure 3.
(a) Seventy three year-old male with paraseptal emphysema. Axial CT images show subpleural (long arrows) and parafissural (short arrows) distribution of the emphysematous spaces demarcated by thin interlobular septa. (b) Different patient, 57 year-old male with paraseptal emphysema involving peri-bronchovascular interstitium (arrowheads).

Figure 4.
Forty one year-old male with vanishing lung syndrome and recurrent pneumothorax presenting with sudden onset shortness of breath. Axial (a) and coronal (b) images of a CT demonstrated left basilar pneumothorax and large bullae, greater in the left lung. Note collapsed left lower lobe simulating a mass (arrow in a). The bullae in the left upper lobe occupy more than one third of the hemithorax compatible with VLS.

Panacinar or panlobular emphysema typically affects the lower lung zones and is usually seen in alpha-1 antitrypsin deficiency. Other causes of panlobular emphysema include Swyer-James syndrome and intravenous injection of talc-containing substances (usually in the context of illicit drug abuse). Pathologically, the entire acini, from respiratory bronchioles to alveoli, are uniformly enlarged [3, 4]. On CT, this appears as extensive areas of uniform low attenuation without sparing of any component of the secondary pulmonary lobule or acinus. This commonly results in a greater degree of lung inflation than in centrilobular emphysema [3] (Figure 5).

Figure 5.
Sixty four year-old female with alpha-1 antitrypsin deficiency and classical radiographic manifestations of panlobular emphysema. (a) PA CXR shows markedly increased pulmonary lucency and decreased vascularity, flattening of the diaphragm (dotted line) and widening of the intercostal spaces. (b) Axial CT image demonstrates marked hyperinflation and paucity of pulmonary vessels of the bilateral lower and right middle lobes. Note atelectatic band in the right lower lobe (arrow), likely a result of compression by hyperinflated lung.

Combined pulmonary fibrosis and emphysema syndrome (CPFE) is characterized by simultaneous occurrence of emphysema and pulmonary fibrosis. The latter commonly being of usual interstitial pneumonia (UIP) – type pattern. Spirometry testing in CPFE usually demonstrates preservation of lung volumes and normal FEV1/FVC, due to the neutralizing effects of coexisting obstructive and restrictive physiologies. On the other hand, the combined effect of emphysema and fibrosis results in a severely impaired gas exchange diagnosed as low diffusion capacity of the lung for carbon monoxide (DLCO), in these patients. The exact pathogenesis of CPFE remains uncertain, however a compiled review of published cases showed that 98% of patients with CPFE were smokers or former smokers and 90% were men [6]. Although prior reports counter-intuitively demonstrated better survival in patients with CPFE than in patients with IPF alone, Ryerson et al. recently showed no difference in survival between the two groups [7].

Imaging of CPFE is characterized by upper lobe emphysema and lower lobe pulmonary fibrosis. The emphysema may be bullous, paraseptal or centrilobular [6]. The most common fibrosis pattern in CPFE is usual interstitial pneumonia (UIP), with basilar predominant reticulations, traction bronchiectasis and bronchiolectasis, and honeycombing. Non-specific interstitial pneumonia (NSIP)-type pattern, and ground glass changes suggesting RB-ILD or DIP have been reported as well [8]. Variability of the fibrosis pattern and severity in CPFE may partly explain the conflicting reports of better survival in patients with CPFE compared to IPF. Pulmonary hypertension is an important complication and cause of mortality in CPFE and can be suggested by the presence of an enlarged pulmonary trunk on imaging. Lung cancer seems to be more prevalent in patients with CPFE than in those with isolated COPD or idiopathic pulmonary fibrosis (IPF) [6].

Giant bullous emphysema/vanishing lung syndrome (VLS) is characterized radiographically by the presence of unilateral or bilateral upper lobe giant bullae occupying at least one third of the hemithorax with associated compression of adjacent normal lung [9] (Figure 4). It is predominantly seen in young male smokers, who present with worsening dyspnea due to progressive expansion of the bullae [10].

On chest radiographs VLS is characterized by unilateral or bilateral thin walled lucencies in the upper lobes. Adjacent band opacities may be seen and represent adjacent compressed lung. The findings on CT include variable number of bullae varying in size between 1 and 20 cm. CT also frequently shows associated paraseptal emphysema and, to a lesser extent, centrilobular emphysema [11]. Imaging can also reveal complications of VLS, such as pneumothorax resulting from ruptured bullae, infected bullae and bronchopleural fistulas [10]. It can be difficult to distinguish the displaced visceral pleural line of pneumothorax from the linear bands seen in VLS on CXR, and CT may be needed for confirmation. In symptomatic patients, bullectomy is the treatment of choice and preoperative determination of the volume of the bullae by CT can predict the degree of improvement in lung function and FEV1 following surgery [10].

Non smoking-related emphysema: In 10% of patients, causes other than smoking are believed to result in the development of emphysema [12]. As with smoking-related emphysema, these may or may not be associated with persistent airflow limitation and therefore may or may not fit the definition of COPD. The morphological pattern of emphysema in non-smokers is sometimes vague or not specified in the medical literature, which may be partly due to its limited documentation on chest radiographs. The most common causes in adults include.

Alpha 1-Antitrypsin deficiency (AAT) is an autosomal co-dominant disease characterized by reduced levels of serum alpha-1 antitrypsin. This deficiency results in unopposed neutrophil elastase action. This results in lung destruction, hepatic cirrhosis and panniculitis. Severity of the disease is variable and dependent on specific genotype. It is often detected between the age of 35 and 45. Treatment includes substitution by exogenous alpha-1 antitrypsin [12].

On imaging ATT manifests as basal predominant panlobular emphysema (Figure 5). Mild disease is very difficult to distinguish from normal lung on CXR or even CT. CXR demonstrates increased lucency and reduced vascularity at the lung bases and signs of hyperinflation. CT shows widespread areas of decreased attenuation without definable walls, reflecting panlobular emphysema at the bases, and normal upper lobes. Other CT findings include paucity of the vessels in the affected areas, signs of hyperinflation such as intercostal bulging of the lungs and a saber sheath trachea. Several studies also report the presence of bronchiectasis and bronchial wall thickening in affected areas of the lung, possibly due to recurrent infection [13]. Smoking can accelerate AAT progression and associated decline in lung function [12].

Connective tissue disorders. Emphysema can be seen in patients with connective tissue disorders, including Marfan, Ehler-Danlos and Cutis laxa syndromes, thought to result from a defect in elastin. In Marfan syndrome emphysema is commonly paraseptal [14] and there may be formation of apical bullae, leading to pneumothoraces (Figure 6) [15]. To our knowledge, there is a single report of panacinar emphysema in Ehler-Danlos [16]. Although there are several reports of emphysema in cutis laxa, the specific morphological type is not specified in these publications, but described as ‘bullous’ [17].

Figure 6.
Twenty eight year-old male with Marfan syndrome. (a) PA CXR shows extensive bilateral upper lung lucencies and absence of bronchovascular markings consistent with large bullae. There is compressive atelectasis, best appreciated in the right lower lung, from the mass effect of the large bullae. (b) Coronal oblique CT confirms these findings. The upper lobes are almost entirely replaced by large bullae and there is compression of the lower lungs. The patient subsequently underwent bilateral lung transplantation.

Cicatricial emphysema, also known as paracicatricial or irregular emphysema represents airspace enlargement and lung destruction regardless of the boundaries of the acinus or secondary lobule and is usually the result of adjacent areas of scarring or fibrosis in inflammatory and granulomatous processes, such as pulmonary progressive massive fibrosis [3, 18] (Figure 7). Since this type of emphysema is always associated with fibrosis, it does not strictly meet the criteria for the definition of COPD, and is usually not associated with obstructive physiology [3].

Figure 7.
Fifty six year-old male with history of crack cocaine smoking, cigarette smoking and IV drug abuse. Coronal CT image demonstrates a large calcified mass consistent with progressive massive fibrosis. There is adjacent lung destruction resulting in irregular/paracicatricial emphysema (long arrow). Note marked circumferential bronchial wall thickening (short arrow) stable for 6 years, consistent with chronic bronchitis.

Emphysema related to illicit drug abuse. Emphysema and bullous lung disease have been reported in several cases of intravenous (IV) injections of talc filler-containing drugs intended for oral administration (e.g. methylphenidate and methadone tablets), and in heroin and cocaine smokers.

In IV users of talc-containing substances, especially of methylphenidate, the emphysema pattern is panlobular, basilar and symmetric, indistinguishable for alpha-1 antitrypsin deficiency, unless associated with findings of progressive massive fibrosis in the upper lungs, related to talcosis [19, 20] (Figure 7).

Upper lobe bullous emphysema has been described in heroin and cocaine smokers [21, 22]. Despite being the world’s most widely used illicit drug, there are only case reports describing upper lobe bullous emphysema in cannabis (marijuana) smokers [23]. In view of recent legalization of cannabis in many US states and countries, and potential increased recreational use, systematic analytical studies are needed to assess its association with emphysema.

Emphysema in HIV patients. HIV seropositivity is an independent risk factor for the development of COPD and emphysema [24]. Diaz et al. found a 15% incidence of emphysema in a cohort of HIV positive patients compared to a 2% incidence in the control HIV-negative group [24]. The mechanism is not fully understood and could be potentially the result of complex immune, apoptotic, proteolytic and oxidative stress responses causing lung destruction over time [25]. On imaging, emphysema preferentially affects the upper lobes and should not be confused with cystic lung disease seen in Pneumocystis jirovecii infection.

Hypocomplementemic Urticarial Vasculitis syndrome (HUVS) is a rare immune complex-mediated disorder affecting multiple organs and clinically presenting with chronic urticaria, angioedema, arthritis or synovitis, conjunctivitis, uveitis, renal insufficiency and abdominal pain. It is 8 times more common in women, usually peaks in the 5th decade of life and may be associated with systemic lupus. Diagnosis is made by skin biopsy showing leukocytoclastic vasculitis and/or presence of anti-C1q antibodies in serum. HUVS may result in panlobular basilar emphysema indistinguishable from that of alpha-1 antitrypsin deficiency. Lung involvement is an important cause of morbidity and mortality in HUVS patients [26].

3. Radiography

Chest radiography plays a frontline role in the assessment of COPD and COPD exacerbations. It has the advantage of being a fast and easy exam to perform in various settings, requires minimal cooperation from patients, has minimal radiation exposure and is inexpensive.

The radiographic signs of emphysema and accuracy of its diagnosis by chest radiography were comprehensively assessed in the classic study by Thurlbeck and Simon [27], who correlated pathological findings of emphysema in 696 necropsies with ante mortem radiographs and pulmonary function test data. The salient findings of this study were five-fold. (1). The diagnostic accuracy of emphysema is highly dependent on the severity of the disease: the frequency of accurate radiological diagnosis was less than 5% in mild disease, 12–20% in moderate disease, and 50%–67% in cases of severe emphysema. (2). Corroborative clinical history (e.g. known alpha-1 antitrypsin deficiency or history of unilateral lung transplant) increases diagnostic accuracy. (3). Panacinar emphysema is more likely to be diagnosed radiographically than other types of emphysema (Figure 5). (4). Superimposed acute or chronic lung disease might obscure the typical radiographic findings of emphysema: the frequency of accurate diagnosis of emphysema dropped from 67 to 41% in patients with a superimposed active lung disease and (5). Quantitative radiographic assessment of emphysema is of little benefit and in clinical practice the diagnosis of emphysema on chest radiograph remains qualitative and highly subjective. The interpretation of findings is prone to inter-observer and intra-observer variability.

The radiographic features of emphysema reflect the presence of parenchymal destruction, vascular remodeling and hyperinflation (Table 2). These comprise the following:

Signs of parenchymal destruction and vascular remodeling

→Bullae

→Increased radiolucency of the lungs

→Vascular attenuation or complete absence of vasculature

Signs of hyperinflation and air trapping

→Flattening of the diaphragm

→Increased retrosternal space

→Increased AP diameter of the chest

→Widened intercostal spaces

→Narrowed cardiac silhouette

Supportive signs

→Saber sheath trachea

→Increased lung markings

Table 2.

Radiographic features of emphysema.

Bullae are the result of coalescing emphysematous spaces, 1 cm or more in diameter [28]. They are the only direct sign of emphysema, sometimes seen on chest radiographs but better appreciated on CT. Bullae appear as well-demarcated areas of increased lucency, devoid of lung markings (Figure 6).

Increased radiolucency of the lungs is the result of small, scattered emphysematous spaces and constitutes an indirect sign of lung destruction (Figures 1,5,6, and 8).

Figure 8.
Sixty seven year-old male, ex-smoker with 100 pack-years smoking history, with centrilobular emphysema and COPD. Coronal CT demonstrate COPD quantification using commercially available software (Philips IntelliSpace portal, Philips healthcare V8.0). Using a threshold of −950 HU it shows centrilobular emphysema present in all lobes (red color). Volumetric measurements and emphysema ratios are obtained for both lungs and all lobes. The emphysema ratio for both lungs is 8.5%. Emphysema is worst in the right upper lobe (emphysema ratio 11.3%).

Vascular attenuation is present when there is paucity or absence of vessels in the outer third of the lung (Figures 1,5, and 9). Vessels may be completely absent secondary to bullae or in severe emphysema, or they can be present but narrowed or impoverished (“pruning”). The causes of this phenomenon may include passive compression of small vessels in the lung periphery by emphysematous spaces, hypoxic vasoconstriction and/or vascular remodeling [29].

Figure 9.
Fifty year-old female with end-stage COPD, undergoing pre-transplant assessment. (a) frontal CXR shows increased radiolucency and vascular attenuation/irregularity especially in the upper lungs, widened intercostal spaces and elongation of the cardiac silhouette. (b) Lateral CXR shows increased retrosternal space, increased AP diameter of the chest and flattening of the hemidiaphragms. Notice enlarged pulmonary artery in this patient with associated pulmonary hypertension. (c) axial CT image confirms marked increased radiolucency (−990 HU in the right upper lobe) and vascular irregularity (arrows) predominating in the upper lungs, compatible with severe emphysema. Coronal (30° caudal angulation) CT ventilation map (d) showing abnormally low attenuation values in the upper lungs (close to -1000HU) compatible with severe emphysema. The patient subsequently underwent double lung transplantation.

Hyperinflation results from air-trapping and bullous formation, and manifests itself radiographically as diaphragmatic flattening, increased retrosternal space and antero-posterior (AP) diameter of the chest on lateral radiographs (“barrel chest”), widened intercostal spaces and narrowed cardiac silhouette on frontal radiographs (Figures 1 and 9).

Saber sheath trachea refers to the radiographic appearance of the trachea when its sagittal-to-coronal diameters ratio is greater than 2. This tracheal configuration is highly associated with obstructive airway disease [30]. It may be the result of chronic airflow limitation during expiration [31].

Increased lung markings also referred to as “dirty lungs” are the result of chronic bronchitis, which is often concomitant with emphysema.

When considered separately, each of the above radiographic features has low specificity and sensitivity for diagnosis of emphysema. However, co-occurrence of several of these findings allows detection of most cases of moderate and severe emphysema and some cases of mild emphysema [32]. In a study by Sutinen et al., the combined findings of hyperinflation and vascular alterations allowed accurate diagnosis of emphysema in 97% symptomatic and in 47% of asymptomatic subjects with necropsy-proven emphysema [33].

4. Computed tomography (CT)

CT is superior to chest radiography in detection of emphysema. Similarly, disease progression is much easier to detect on CT than on chest radiography. CT also provides superior information regarding distribution and extent of the disease, which might allow determination of the cause of emphysema and its subtype.

Emphysema can be readily detected on conventional CT with 5–10 mm slice thickness, however is best depicted with high-resolution CT with 1–2 mm slice thickness reconstructed with a lung algorithm [34]. Using modern multislice CT scanners, such acquisitions can be obtained at full inspiration during a single breath-hold. Coronal and sagittal reformations are helpful in assessment of the distribution and extent of emphysema. Although intravenous contrast is not required for the assessment of emphysema, pulmonary blood volume imaging obtained from dual energy CT (a computed tomography technique that uses two separate x-ray photon energy spectra, allowing the interrogation of materials that have different attenuation properties at different energies) may aid in the detection of emphysema by highlighting areas of hypo-perfusion (Figure 10).

Figure 10.
Seventy nine year-old female with a 50 pack-year history of smoking. Axial CT (a) and corresponding pulmonary blood volume image (b) at midlung level show decreased perfusion in areas of emphysema.

On CT, emphysema appears as focal, regional, or diffuse areas of low attenuation, contrasting with surrounding normal lung. In addition, all conventional radiographic features of emphysema, such as the attenuation of the caliber of the vessels in emphysematous regions, may be identified on CT. Features of hyperinflation may be better appreciated on coronal and sagittal reformations than on axial images. The severity of emphysema in various regions of the lungs can be visually and subjectively estimated by scrolling though all available CT images in any plane and rating them on a four point scale as (1) < 25% of the area (2) 26%–50% of the area, (3) 51%–75% of the area, or (4) > 75% of the area [31].

When correlated with histopathology (gold standard), the accuracy of CT for both, detection of emphysema and of its distribution, increases with thin collimation high-resolution computed tomography (HRCT) [35]. When correlated with the pathologic grade of emphysema, HRCT performance is excellent in vitro (r = 0.91), and slightly lower but strong-to-very strong in vivo (r = 0.7–0.9), although very mild emphysema may be missed [32].

CT is more sensitive than pulmonary function tests for the detection of emphysema. In a study involving 615 men ranging in age from 40 to 69 years, who underwent lung cancer screening with low-dose spiral CT, emphysema was detected in 30% of current smokers (116/380); of these, the majority had normal spirometry results (78%). Additional studies have also shown that 68–80% of smokers with emphysema detected on HRCT had normal spirometry results [35].

The sensitivity for detection of subtle emphysema can be improved by using the MinIP (minimum intensity projection) technique [36]. Contrary to the MIP (maximum intensity projection), which is helpful in the detection of high attenuation structures such as vessels and lung nodules, the MinIP technique recognizes the regions of the lung with the lowest attenuation values such as emphysema, while subtracting the normal lung and pulmonary vasculature. This technique is however not widely used in clinical practice due the extra steps involved in producing these images and their limited clinical value.

5. Quantitative assessment of emphysema

The two main techniques used to quantify the extent of emphysema on CT are the threshold technique and histogram analysis. These are based on the simple fact that emphysematous lung has lower attenuation than normal lung parenchyma. Since the introduction of these techniques several variations have been introduced, mainly as a result of improvements in CT capabilities and image processing software. Such examples include 3D assessment of whole lung density, made possible by fast image acquisition of the entire lung volume using multidetector CT.

Threshold technique. In this technique, emphysema is considered to be present if the attenuation value of the pixels in the area of interest falls below an absolute predetermined Hounsfield unit threshold value [37]. The threshold value of – 950HU for detection of emphysema has been shown to correlate well with pulmonary function tests and pathological data when using thin section CT at 10 mm interval (Figure 8) [38, 39]. (For reference, the value of −1000 HU corresponds to radiodensity air and 0 HU to that of water.) Prior to that, Muller et al. had used a software program called density mask to highlight voxels falling within a predetermined range, and found that a threshold value of −910 HU correlated best with the extent of emphysema when using a 10 mm slice thickness [40]. The use of other threshold values, ranging between −900 and −980 HU, has also been reported and is dependent on a variety of factors including scanning parameters [37].

Histogram analysis. In this technique, emphysema is detected if the attenuation value of a pixel falls below a predetermined percentile. The enlargement of the air spaces distal to the terminal bronchioles, accompanied by destruction of alveolar walls in emphysema, results in a reduced ratio of the surface area of the walls of distal airspaces per unit lung volume (AWUV) [41]. In their CT-pathologic correlation study, Gould et al. showed significant correlation between AWUV and the lowest 5th percentile of the CT density histogram (r = −0.77) [41]. Contrary to the threshold technique, where an absolute number is predetermined as an indicator of emphysema, histogram analysis can underestimate the extent of emphysema if a concomitant disease (such as pulmonary mass or consolidation) shifts the histogram curve towards higher overall Hounsfield values [34].

6. Imaging of comorbidities associated with COPD

Imaging is also valuable in establishing the presence of co-morbidities frequently associated with COPD. The most common association is with heart failure and ischemic heart disease. Emphysema can also eventually lead to pulmonary hypertension and right heart failure with radiographs and CT showing enlarged central pulmonary arteries and cardiomegaly. Pulmonary hypertension has been estimated to be present in 35 to 90% of patients with COPD and its presence is associated with greater mortality and morbidity [42].

Lung cancer is the most frequent cause of death in patients with mild to moderate COPD, whereas infections are the most common cause of COPD exacerbations and are associated with significant mortality [1].

The radiologist should therefore look for and report signs suggestive of these comorbidities, such as coronary and aortic atherosclerosis, cardiomegaly, enlarged pulmonary arteries, pulmonary edema, pleural effusions, pulmonary nodules, masses and consolidations.

7. Conclusion

Radiological imaging modalities, such as radiography and CT, are crucial in diagnosis and quantification of emphysema, in differentiating among its different subtypes and identification of its potential etiologies, monitoring of disease and complications, and management of these entities. Newer techniques in quantitative CT can provide more objective, reproducible and more reliable longitudinal assessment of emphysema, however these techniques need to be validated in large cohorts and their current use remains limited in clinical practice.

Conflict of interest

The authors declare no conflict of interest.

Ji Lee

Chapters authored

Related collaborators

Valmore Bermúdez

Joselyn Rojas

Miguel Aguirre

Climaco Cano

Carlos Silva

Manuel Velasco

Marcos Lima

Raquel Cano

Nailet Arraiz

Eneida Fonseca