At the completion of this you will be able to:

1) properly use the HRCT scanning descriptive term "Crazy Paving" and state the different disease processes that cause this pattern.

2) properly use the HRCT scanning descriptive term "Mosaic pattern of lung attenuation" and state the different disease processes that cause this pattern.

3) properly use the HRCT scanning descriptive term "Tree-in-bud" and state the different disease processes that cause this pattern.

4) properly use the HRCT scanning descriptive term "Ground glass opacification" and state the different disease processes that cause this pattern.

High resolution CT scanning (HRCT) of the chest often shows findings that produce a very short differential diagnosis, sometimes pathognomonic for certain disease processes, such as centrilobular pulmonary emphysema, bronchiectasis, or classic honeycomb lung fibrosis. Several colorful descriptive terms have been popularized in the radiology literature, becoming HRCT "buzz words". These terms, describing a particular disease process, are sometimes used inappropriately to infer a pathognomonic finding, when in fact, the terms are often non-specific. These HRCT buzz words include "crazy paving", "mosaic perfusion or mosaic pattern", "ground glass opacity", and "tree-inbud".

The term crazy paving has been used to describe the CT scan findings of pulmonary alveolar proteinosis. The terms mosaic perfusion or "mosaic oligemia" describe the chest CT scan finding of pulmonary thromboembolism. The term ground glass opacity, or ground glass attenuation, has been used for many years to describe any radiographic or CT scan finding of a hazy increase of opacity involving the lungs, as well as bones in fibrous dysplasia, and was popularized as a chest HRCT scan finding of alveolitis in patients with usual interstitial pneumonia . The term tree-in-bud has also been used for many years, dating back to the bronchogram descriptions of respiratory  bronchioles by Twining and Kerley, but more recently popularized as a chest CT scan finding of active endobronchial tuberculosis. These terms and the findings they represent are much less specific than once thought and are now recognized to be associated with a number of disorders. In this perspective, we will discuss these HRCT chest buzz words and review the proper use of the terms and their significance.

Crazy paving is a colorful descriptive term for the HRCT scan findings of apparent thickened interlobular septa and intralobular structures, in areas of ground glass opacity, forming typical polygonal shapes, with no architectural distortion. The diseased lung is usually quite well demarcated from surrounding normal lung tissue, creating a "geographic" pattern. Crazy paving is thought by many to be a pathognomonic finding of pulmonary alveolar proteinosis. The finding is still strongly suggestive of pulmonary alveolar proteinosis in the appropriate clinical setting. However, in isolation, this HRCT scan pattern has also been described with such diverse disease processes as Pneumocystis carinii pneumonia, sarcoidosis, bronchioloalveolar carcinoma, ARDS, and exogenous lipoid pneumonia.

The term "mosaic perfusion" was popularized by Martin et al as a CT scan finding of pulmonary thromboembolic disease resulting from regions of hyperemic (higher attenuation) lung adjacent to oligemic (lower attenuation) regions of lung. While acute pulmonary thromboembolism usually does not produce a mosaic pattern of lung attenuation, similar CT findings of mosaic perfusion are described for pulmonary arterial hypertension. In addition to vascular disease, a mosaic pattern of lung attenuation has been described in two other disease categories: (1) small airways disease, (2) infiltrative lung diseases. Areas of variable lung attenuation in a lobular or multilobular distribution are almost never a normal finding, except as an easily distinguished normal gravitational gradients of lung density. The mosaic pattern of lung attenuation presents a challenge to the radiologist when deciding which are the abnormal regions of lung, those of low attenuation or high attenuation, or both. It is often possible to distinguish among these categories by using the following additional CT scan findings. CT performed at suspended full expiration shows the physiologic consequence of small airways (bronchiolar) diseases: air trapping. Lung regions that retain air during exhalation remain more lucent and show less decrease in volume than lung supplied by normal airways. The distribution of air trapping is often patchy and dependent on the level and severity of the airway obstruction. When the level of airway obstruction is at the subsegmental or lobular level, a mosaic pattern of normal lung and hyperlucent lung can result. Lung regions that retain air show a decrease in the caliber and number of pulmonary vessels relative to normal lung. The inciting pathologic processes can be permanent, such as seen in patients with obliterative bronchiolitis or reversible, such as seen in patients with asthma. In some instances, air trapping can be completely unsuspected on routine suspended full inspiration CT scanning and only become evident on CT scans obtained at suspended full expiration. In small airways diseases, the lucent regions of lung seen at inspiration remain lucent at expiration because of air trapping, showing no or minimal increase in lung attenuation and no or minimal decrease in volume. In chronic vascular disease, because there is no air trapping or airway disease, the attenuation of both the hyperemic and oligemic lung at inspiration will increase in a similar fashion and the volume of both will decrease uniformly at expiration. A patchy infiltrative process within the interstitium of the lung or partial filling of the air spaces by fluid, cells, or fibrosis can occur such that the CT attenuation of the affected lung increases relative to that of normal parenchyma. This patchy distribution can appear as a typical mosaic pattern. The vessel caliber and number are not appreciably different between the normal and abnormal regions of lung. Diseases that can produce such a CT pattern of mosaic lung attenuation include Pneumocystis carinii pneumonia, chronic eosinophilic pneumonia, hypersensitivity pneumonia, bronchiolitis obliterans organizing pneumonia, or bacterial pneumonia.

The term tree-in-bud has also been used for many years, dating back to the bronchogram descriptions of respiratory bronchioles by Twining and Kerley, but more recently popularized as a chest CT scan finding of active endobronchial tuberculosis. The "tree-in-bud" (TIB) pattern is a direct CT scan finding of bronchiolar disease. Gruden and colleagues have described this same appearance as resembling the childhood toy "jacks". This pattern is analogous to the larger airway "finger-in-glove" appearance of bronchial impaction, but on a much smaller scale. The TIB pattern has become a popular descriptive term for many bronchiolar disease processes, all with similar appearances, though is still often used inappropriately to imply a pathognomonic finding for tuberculosis. The list of diseases associated with the bronchioles potentially producing a TIB pattern at CT scanning is extensive. The more common disease processes can be grouped as follows:

1) infection,

2) immunologic disorders, such as allergic bronchopulmonary aspergillosis,

3) congenital disorders, such as Kartagener’s syndrome,

4) aspiration, and

Aspiration of infected oral secretions or other material is the most common cause of the treein-bud appearance in our experience. Endobronchial spread of Mycobacterium tuberculosis represents a chronic granulomatous infection in which active organisms spread via the airways after necrosis of a bronchial wall and softening or liquefaction necrosis of caseous material. When the spread is pathologically extensive enough to be detected with CT scanning, the earliest CT scan finding is the TIB pattern of 2 - 4 mm centrilobular nodules and branching linear structures of similar caliber originating from a single stalk. This TIB appearance is characteristic, but not pathognomonic, of active, likely contagious, tuberculosis. In the proper clinical setting, the TIB pattern is thought to be a reliable criterion for disease activity, distinct from old fibrotic lesions. A typical example is an upper lobe cavitary lesion containing active infectious organisms that subsequently "spills" out into lower lung airways, resulting in bronchogenic spread of disease. Indirect CT scan signs of bronchiolar disease include air-trapping, especially with expiratory CT scanning, and subsegmental atelectasis.

Ground glass opacification describes a finding on high resolution CT of the lungs in which there is a "hazy increased attenuation of lung, with preservation of bronchial and vascular margins; caused by partial filling of air spaces, interstitial thickening, partial collapse of alveoli, normal expiration, or increased capillary blood volume; not to be confused with consolidation, in which bronchovascular margins are obscured; may be associated with an air bronchogram". This finding is often occult on chest radiographs. GGO can represent either normal or abnormal interstitial or alveolar processes, findings beyond the resolution of the HRCT technique. The term ground glass opacity, or ground glass attenuation, has been used for many years to describe any radiographic or CT scan finding of a hazy increase of opacity involving the lungs, as well as bones in fibrous dysplasia, and was popularized as a chest HRCT scan finding of alveolitis in patients with usual interstitial pneumonia. Ground glass opacification has become a non-specific finding, but which in certain clinical circumstances can suggest a specific diagnosis, indicate a potentially treatable disease, or guide a bronchoscopist or surgeon to an appropriate area for biopsy. GGO is a frequent finding on HRCT, with a lengthy differential diagnosis. It is very important to correlate the HRCT scan finding of GGO with the clinical presentation to narrow the lengthy differential diagnosis. The following disease processes can all result in a ground glass pattern as the sole, dominant, or accompanying HRCT scan manifestation: lung parenchymal fibrosis with or without active inflammation, pulmonary contusion, acute pulmonary hemorrhage from any etiology, bronchioloalveolar carcinoma, intrathoracic lymphoproliferative disorders, cardiac and noncardiac pulmonary edema, adult respiratory distress syndrome, sarcoidosis, and infectious pneumonia of any etiology. Hypersensitivity pneumonitis is an interesting disease process that can manifest several of the HRCT "buzzwords" including a mosaic pattern of lung attenuation, GGO, and of air trapping. Hypersensitivity pneumonitis is a complex immunologic reaction by the lung, primarily to inhaled organic antigens, although non-inhaled drugs can also be an inciting agent. The most important and common disorders and their inciting agents are farmer’s lung from thermophilic actinomycetes in moldy hay, and bird fancier’s lung from avian protein in droppings and feathers. The clinical presentation may be categorized as acute, subacute, or chronic, depending on the periodicity of exposure and quantity of inhaled antigen. GGO is a dominant feature in subacute hypersensitivity pneumonitis. There are even some non-disease states—normal conditions, and technical factors—producing GGO. For example, narrow window widths and levels can erroneously create the appearance of GGO. Thick (5- 10 mm) collimation may cause a false appearance of GGO that is shown to be volume averaging of linear structures with thin sections. Subpleural opacities are frequently identified in the dependent portions of the lungs, as a result of gravity and microatelectasis. These dependent opacities consist of reticular, linear, and ground glass opacities, that can be confused with or mask true infiltrative lung disease. Repeat scanning with the patient in the prone position can make the distinction between infiltrative lung disease and nonpathologic dependent opacities. Lung attenuation normally increases with exhalation. This increased attenuation can mask underlying GGO, or create an appearance of diffuse lung disease if the expiratory nature of the examination is not appreciated.

1. Franquet T, Gimenez A, Bordes R, Rodriguez AJ, Castella J. The crazy-paving pattern in exogenous lipoid pneumonia: CT-pathologic correlation. AJR 1998; 170:315-317

2. Murch C, Carr D. Computed Tomography Appearances of Pulmonary Alveolar Proteinosis. Clinical Radiology 1989; 40:240-24

1. Martin KW, Sagel SS, Siegel BA. Mosaic oligemia simulating pulmonary infiltrates on CT. AJR 1986; 147:670-673

2. Stern EJ, Swensen SJ, Hartman TE, Frank MS. CT mosaic pattern of lung attenuation: distinguishing different causes. AJR 1995; 165:813-816

3. Stern EJ, Müller NL, Swensen SJ, Hartman TE. CT mosaic pattern of lung attenuation: etiologies and terminology. J Thorac Imaging 1995; 10:294-297

4. Arakawa H, Webb WR, McCowin M, Katsou G, Lee KN, Seitz RF. Inhomogeneous lung attenuation at thin-section CT: diagnostic value of expiratory scans. Radiology 1998; 206:89-94

1. Gruden JF, Webb WR. Identification and evaluation of centrilobular opacities on high-resolution CT. Semin Ultrasound Ct Mr 1995; 16:435-449

2. Im JG, Itoh H, Shim YS, et al. Pulmonary tuberculosis: CT findings—early active disease and sequential change with antituberculous therapy. Radiology 1993; 186:653-660

3. Collins J, Blankenbaker D, Stern EJ. CT patterns of bronchiolar disease: what is "tree-in-bud"? AJR 1998; 171:365-370

1. Remy JM, Remy J, Giraud F, Wattinne L, Gosselin B. Computed tomography assessment of ground-glass opacity: semiology and significance. J Thorac Imaging 1993; 8:249-264

2. Collins J, Stern EJ. Ground glass opacities on CT scan of the Chest—an ABC approach to diagnosis AJR 1997;169:355-367