Faculty Bibliography

The potential of low-dose computed tomography (CT) of the lungs was critically evaluated in two patients with normal-appearing lungs and 10 patients with a wide diversity of underlying parenchymal abnormalities. At each of five levels, in addition to routine scans obtained at 120 kVp and 140 mA, a scan at 10 mA and a half scan at 10 mA were obtained, with all other parameters held constant. Each scan was evaluated visually to assess anatomic clarity as well as the presence of artifacts and the extent of graininess. At all levels of the thorax, visualization of parenchymal structures was not affected by decreasing the milliamperage. It appears that high-quality, diagnostic images of the lung can be obtained with a very low radiation dose. Although further evaluation is necessary, the potential of low-dose CT for use in the pediatric population in particular, as well as for screening in patients at high risk for developing lung cancer, is apparent.

In patients with tumors of tracheobronchial origin, computed tomography (CT) provides detailed information critical to accurate tumor staging. By precisely delineating the intra- and extraluminal extent of disease, CT has proved complementary to fiberoptic bronchoscopy, both for initial screening and as an aid in determining appropriate candidates for either transbronchial needle aspiration and biopsy or endoscopic laser therapy. CT has proved especially valuable as a guide to patients with radiographic evidence of lobar or segmental collapse. The advantages and limitations of CT as well as magnetic resonance (MR) in the evaluation of tracheobronchial neoplasia are discussed and illustrated.

Lung hernia is an abnormal protrusion of the lung beyond the confines of the thoracic cage. Hernias are classified as cervical, intercostal, or diaphragmatic. Each of these types can be either congenital or acquired. Acquired hernias can be spontaneous, posttraumatic, or pathologic as a result of a neoplastic or inflammatory process. This report describes five cases of lung hernias. Two were congenital cervical hernias, detected as incidental findings on chest radiographs of infants. One was a posttraumatic intercostal hernia detected on a chest radiograph and confirmed with CT. The remaining two were acquired intercostal hernias at the site of prior chest-tube drainage of tuberculous empyemas. Although both of these were clearly shown on CT scans, only one was visible on chest radiographs.

Recent reports have emphasized the potential for dual-energy computed radiographic applications. An improved method for single-exposure material-selective imaging with a photostimulable phosphor computed radiography system was investigated. The essential elements of the technique are (a) prefiltration with gadolinium, which divides the incident broad-beam x-ray spectrum into low-energy and high-energy peaks; (b) a cassette consisting of four photostimulable phosphor plates that record images of increasing mean energies, with a computed energy separation of 23 keV from the front to the rear plate; (c) spatially dependent scatter and beam-hardening corrections; and (d) a noise-reduction algorithm based on noise correlations between bone-selective and soft-tissue-selective dual-energy images. These elements result in improved material cancellation and signal-to-noise ratio throughout the image.

The role of CT in the diagnosis of intrathoracic Kaposi's sarcoma (KS) was evaluated retrospectively in 24 patients, in the absence of coexistent opportunistic infections. In all cases the diagnosis of KS was initially established by histologic evaluation of extrathoracic disease: 15 patients had verified parenchymal KS and nine patients endobronchial KS. (Chest roentgenograms were analyzed separately for each group: in 14 patients serial films were available for review. The predominant radiographic findings was the presence of nonspecific, bilateral, perihilar infiltrates in 22 of 24 cases (92 percent). Corresponding CT scans documented the presence of abnormal hilar densities characteristically extending into the adjacent pulmonary parenchyma along distinctly perivascular and peribronchial pathways. Discrete, poorly marginated nodules were identified radiographically in ten cases (42 percent); these proved to be randomly distributed throughout the parenchyma on CT. Radiographic evidence of mediastinal adenopathy was distinctly unusual, seen in only two cases (8 percent). While CT typically demonstrated shotty adenopathy, significantly enlarged nodes (greater than 1 cm) were rarely identified. We concluded that CT is more specific than routine roentgenograms for identifying pulmonary KS. While not pathognomonic, peribronchial and perivascular disease is sufficiently characteristic to obviate more invasive diagnostic procedures, especially in patients with established KS

Radiographic and CT findings in 15 patients with clinically documented septic pulmonary emboli were compared retrospectively. In most cases, radiographic changes were nonspecific. In comparison, on CT a combination of specific signs could be identified in all patients. These included peripheral nodules with clearly identifiable feeding vessels associated with metastatic lung abscesses (10 [67%] of 15 cases), and subpleural, wedge-shaped densities with and without necrosis caused by septic infarcts (11 [73%] of 15 cases). Ancillary pleural, mediastinal, axillary, and pericardial abnormalities also were more easily identified with CT. We conclude that CT is complementary to other imaging techniques in the recognition of septic pulmonary emboli

The authors retrospectively evaluated radiographs, computed tomographic (CT) scans, and results of pulmonary function tests (when available) for 17 patients with biopsy-proved pulmonary histiocytosis X. In 11 patients, high-resolution CT was used. In 12 patients, CT demonstrated cystic air spaces, usually less than 10 mm in diameter. In three of these 12, cysts were the only abnormality, but in six others, nodules (usually less than 5 mm in diameter) were also present. Two patients had only nodules and one, only emphysema. CT showed that many lesions that appeared reticular on plain radiographs were actually cysts. CT showed no central or peripheral concentration of lesions, but it did reveal that many small nodules were distributed in the centers of secondary lobules around small airways. CT findings correlated better with the diffusing capacity (rho = -0.71) than did the plain radiographic findings (rho = -0.57). Thus, CT was better than radiography at showing the morphology and distribution of lung abnormalities.

Thin (1.5- and 5.0-mm) section contiguous computed tomographic (CT) scans obtained through the basilar segmental bronchi in 31 patients were reviewed in order to delineate normal anatomy and common variations of lower lobe airways. In each case, the frequency with which individual segmental and subsegmental bronchi were seen was established, as were variations in branching patterns. All basilar segmental bronchi were identified except in one case in which images of the left lung were obscured due to respiratory and cardiac motion. In the right lung, a division into subsegmental bronchi was identified in 84 of 150 (56%) visualized segmental bronchi. Six separate patterns of basilar segmental subdivision were found. In the left lung, subsegmental bronchi were identified arising from 51 of 145 (35%) visualized segmental bronchi. Five separate patterns of bronchial subdivision were found in the left lung. It is concluded that thin-section CT allows precise identification of all basilar segmental bronchi and, consequently, can play a significant role in the cross-sectional evaluation of lower lobe bronchial and parenchymal abnormalities