The diagnostic value of chest radiography and high-resolution computed tomography (CT) in chronic diffuse interstitial lung disease (CDILD) was assessed in 140 consecutive patients with diffuse infiltration of the lung visible at radiography. Radiographs and CT scans were separately read by three independent observers without knowledge of clinical and pathologic data. The observers listed the three most likely diagnoses and recorded the degree of confidence they had in their choice on a 0%-100% probability scale. Findings at radiography and high-resolution CT were recorded by each observer and were used for a stepwise discriminant analysis between diagnoses. First-choice diagnoses of all three observers that were made with a high level of confidence (probability, greater than or equal to 75%) were more accurate with CT than with radiography (P less than .001). The superiority of high-resolution CT over radiography was most obvious for histiocytosis X and sarcoidosis; in cases of pulmonary fibrosis, CT was not significantly different from radiography. The interobserver agreement for the proposed diagnosis was significantly better with high-resolution CT (P less than .001). Twenty-one of 26 radiographic findings and 21 of 25 CT findings were discriminant. Stepwise discriminant analysis revealed the superiority of CT over radiography, since the ranking of all findings showed that the four most discriminant findings, and eight of the first 12 findings, were revealed with CT.

Forty-four patients with histologically confirmed sarcoidosis were prospectively studied with high-resolution computed tomography (CT). Nodules were seen in all cases. They were isolated in 19 cases and associated with other lesions in 25 cases. Other abnormalities were irregular interfaces (n = 18, 41%), linear network (n = 14, 32%), thickening of the pleural surface (n = 9, 20%), ground-glass opacities (n = 7, 16%), lung distortion (n = 11, 25%), traction bronchiectasis (n = 3, 7%), and network of air-filled cavities (n = 3, 7%). Predominant sites of lesions were the upper and middle zones (n = 30, 68%) and posterior zones (n = 13, 30%). Nodular abnormalities were noted at CT in six cases in which the pulmonary parenchyma appeared normal on radiographs. Lung distortion was noted at CT in eight cases without visible fibrosis on chest radiographs. The majority of patients with lung distortion (nine of 11, 82%) had disease of greater than a years duration. CT improved sensitivity for the detection of all types of lesions, mainly lung distortion. Low but significant correlations were found between visual score at CT and total lung capacity, vital capacity, forced expiratory volume in 1 second, and diffusing capacity.

Eighteen patients with pulmonary lesions of histiocytosis X were studied with high-resolution computed tomography (CT). Thin-walled cysts were found in all but one patient. The other abnormalities included nodules (n = 14), cavitated nodules (n = 3), thick-walled cysts (n = 7), reticulation (n = 4), ground-glass opacities (n = 4), and irregular interfaces (n = 4). The lesions were most often diffuse (n = 16), with a topographic predominance in the upper or middle lung zones in nine patients. Comparison of CT scans and chest radiographs shows that small and large cysts and micronodules are better detected with CT. Comparison of abnormalities found in patients in the early and late stages of the disease, as well as the evolution observed in one case, suggests that CT patterns progress from nodules to cavitated nodules and thick-walled cysts to cysts to confluent cysts.

The purpose of the study was to evaluate the CT abnormalities of airways and lung parenchyma in asthmatic patients and to assess inter- and intraobserver variability for these abnormalities. The CT scans of 50 asthmatic patients and 10 healthy volunteers were assessed independently by four independent chest radiologists who were masked with respect to the clinical information. Bronchiectasis involving mostly subsegmental and distal bronchi was noted in 28.5% of the asthmatic subjects and none of the non-asthmatics. Bronchial wall thickening, small centrilobular opacities and decreased lung attenuation were observed in 82%, 21% and 31% of asthmatic patients respectively, compared with 7%, 5% and 7% of healthy subjects. The intra- and inter-observer agreements for these four CT abnormalities were measured by the kappa statistic and ranged from 0.60 to 0.79 and from 0.40 to 0.64, respectively. It is concluded that asthmatic patients may exhibit bronchial wall thickening, bronchiectasis and morphological abnormalities suggestive of distal airways disease that can be assessed on CT scans with a clinically acceptable observer variability.