757

Determining the Cause of

Pulmonary Atelectasis: A
Comparison of Plain Radiography and CT

John H. Woodring1 In a retrospective analysis of 50 patients with segmental or lobar atelectasis of the
lung, chest radiographs and CT studies were compared for their abilities to distinguish
whether a centrally obstructing tumor was the cause. This was done to help define the
role of CT in evaluating patients with atelectasis. Atelectasis was caused by an obstruct-
American Journal of Roentgenology 1988.150:757-763.

ing tumor in 27 cases and a variety of other conditions in 23. The chest radiograph
correctly identified an obstructing tumor as the cause of atelectasis in 24 of 27 patients
on the basis of the presence of a central hilar mass or obvious bronchial abnormality;
there was 89% sensitivity and 96% specIficity with a 12% false-negative rate and a 4%
false-positive rate. CT correctly identified all 27 obstructing carcinomas on the basis of
the presence of either a central bronchial abnormality or a central hilar mass; there was
100% sensitivity and 87% specificity with a 0% false-negative rate and a 10% false-
positive rate. Absence of air bronchograms or the presence of mucus-filled bronchi
within the atelectatic lung were secondary CT findings that also favored the presence
of an obstructing tumor.
Although the chest radiograph was more specific than CT for tumor as the cause of
atelectasis (96% vs 87%,
respectively), it was less sensitive than CT for tumor (89% vs
100%, respectively) resulting in missed tumor diagnoses. CT identified all cases caused
by obstructing tumor and successfully excluded obstructing tumor in most of the
remaining cases, with an acceptable number of false-positive tumor diagnoses (10%).
CT should be performed when the cause of segmental or lobar atelectasis cannot be
established with certainty on the basis of the chest radiograph.

Patients with segmental or lobar atelectasis are frequently encountered in daily

radiology practice. In most cases, the cause of atelectasis is apparent from the
patient’s history, clinical examination, and chest radiographic findings. However,
the accurate assessment of atelectasis caused by an obstructing tumor is still very
important in initial radiographic evaluation. Although the plain radiographic and CT
manifestations of pulmonary atelectasis are well known, it is unclear how much the
radiologist can rely on the plain film to distinguish atelectasis due to obstruction by
tumor from atelectasis due to other causes. CT has been reliable in showing
endobronchial tumor as a cause of atelectasis [1 ]. However, the sensitivity,
specificity, and false-positive and false-negative rates of plain film and CT diagnoses
seldom have been addressed, and the role of CT in evaluating patients with
atelectasis is uncertain.
In this paper, we review the plain film and CT manifestations of segmental and
lobar atelectasis in order to assess the accuracy of both techniques in distinguishing
Received October 19, 1987: accepted after re-
vision December 22, 1987. obstructing tumor from other causes of atelectasis and to give guidelines for the
Department of Diagnostic Radiology, A. B. use of CT in evaluating atelectasis identified on the initial chest radiograph.
Chandler Medical Center, University of Kentucky,
800 Rose St., Lexington, KY 40536-0084. Address
reprint requests to J. H. Woodnng. Materials and Methods
AJR 150:757-763, April 1988
0361 -803X/88/1 504-0757 The study group consisted of 50 consecutive patients with segmental or lobar atelectasis
© American Roentgen Ray Society of the lung from various causes who had chest radiographs and CT scans performed within
 758 WOODRING AJR:150, April 1988

a 24-hr period. Erect posteroanterior and lateral chest radiographs two cases, a definitive diagnosis of malignancy was established by
were available in 46 patients. These were made at a 6-ft (1 83-cm) CT-guided percutaneous needle biopsy of the hilar mass. In one case
distance using 135 kVp, 300 mA, phototimed exposure, a 1 .2-mm bronchial occlusion from tumor was confirmed at thoracotomy.
focal spot, a three-phase generator, a 1 0:1 focused grid, Kodak Twenty-three patients had other causes of atelectasis, including
Lanex medium-intensifying screens, and Kodak ortho C film for a postoperative or posttrauma atelectasis in seven; pneumonia or other
wide exposure latitude (Eastman Kodak, Rochester, NY). In four acute air-space disease in five; compression caused by pleural effu-
patients, only portable anteroposterior films were available. These sion in five (two had malignant effusion and three had nonmalignant
were made at a 54-in. (1 37-cm) distance using 75 kVp, 5 mAs, a 1.5- effusion, confirmed by thoracentesis and/or thoracotomy); bronchi-
mm focal spot, constant potential output portable generator, no grid, ectasis in three; radiation fibrosis in two; and tuberculous broncho-
Kodak Lanex medium-intensifying screens, and Kodak ortho C film. stenosis in one. In 1 2 of these 23, fiberoptic bronchoscopy had been
The chest radiographs were assessed for adequacy of penetration performed and showed patent bronchi with no endobronchial mass.
of the central bronchi. The trachea was well visualized in all 50, the In nine, the atelectatic segment or lobe reexpanded and remained
right main bronchus in 45, the left main bronchus in 46, and the normal after removal of pleural effusion by thoracentesis or after
proximal lobar bronchi in 21 . Most segmental bronchi were not well appropriate therapy for altered mucociliary transport mechanisms
seen on the plain radiographs. and/or pneumonia. In one patient, normal bronchi were confirmed at
The CT studies were performed on commercially available fourth- thoracotomy, and in one patient with tuberculous bronchostenosis
generation CT scanners. Contiguous i 0-mm collimated scans who was too elderly for bronchoscopy, serial sputum cytologies were
through the bronchi were used routinely; contiguous 5-mm collimated negative and studies for Mycobacterium tuberculosis were positive.
scans were used to evaluate the segmental and lobar bronchi when
the 1 0-mm-thick slices failed to show patency or abnormality of the
bronchi convincingly. A variety of window widths (300 to 2000 H) and
window levels (-30 to -800 H) were used to visualize the lung
Results
parenchyma, bronchi, and mediastinum. All images were obtained in Evaluation of the chest
showed radiographs
that air bron-
American Journal of Roentgenology 1988.150:757-763.

end-inspiration. IV contrast enhancement, accomplished by the hand

chograms were segment
present in the atelectatic
or lobe in
injection of a 50-mi bolus of Reno-M-60 (diatrizoate meglumine injec-
13 of the 50 patients and were absent in 37 (Figs. 1 and 2).
tion USP 60%) (Squibb Diagnostics, New Brunswick, NJ) followed by
Of those with air bronchograms, four (31 %) of 13 had an
the drip infusion of 1 00 ml of Reno-M-60 for 15 mm, was used
routinely. The CT scans were assessed for adequacy of visualization obstructing tumor and nine (69%) did not. In those without
of the proximal bronchi; the trachea, main bronchi, lobar bronchi, and air bronchograms, 23 (62%) of 37 had an obstructing tumor
proximal segmental bronchi were visualized adequately to determine whereas 14 (38%) did not. This suggested that air broncho-
whether they were patent or occluded in all cases. grams had little usefulness in determining the cause of atelec-
The chest radiographs and CT scans were evaluated retrospec- tasis.
tively without knowledge of the cause of atelectasis. For the chest A central hilar mass (S sign of Golden)(Fig. 3) was identified
radiographs the following findings were recorded: segment or lobe in 20 of the 50 cases. All 20 had obstructing tumor; however,
involved, presence or absence of air bronchograms, central hilar
seven of the cases with an obstructing tumor failed to show
mass (S sign of Golden) [2], narrowing or cutoff of the bronchus, and
this sign. None of the other causes of atelectasis showed a
evidence of mediastinal lymphadenopathy. When the central bronchi
were not well seen, bronchial abnormality was not recorded; only
central hilar mass.
when definite narrowing or occlusion could be identified with certainty Definite central bronchial narrowing or cutoff (Fig. 4) was
was this finding recorded. For the CT scans, the following findings identified in 1 8 of 50 cases. In 1 7, obstructing tumor was
were recorded: segment or lobe involved, presence or absence of air present; in one case, tuberculous bronchostenosis was pres-
bronchograms, “mucous” bronchograms, central hilar mass (S sign ent. Ten of the cases with obstructing tumor, however, failed
of Golden), narrowing or cutoff of the bronchus, and evidence of to show a definite central bronchial abnormality.
mediastinal lymphadenopathy. A decision was made as to whether With either a central hilar mass or central bronchial abnor-
the findings were suggestive of atelectasis produced by an obstruct- mality considered as the plain film sign of atelectasis from
ing tumor.
obstructing tumor, sensitivity was 89% (24/27) and specificity
The CT scans had been requested because of an initial clinical
was 96% (22/23). The false-negative rate was 1 2% (3/25),
suspicion of tumor in 31 cases, as part of an evaluation of acute
and the false-positive rate was 4% (1/25). The associated
chest trauma in four, and to help elucidate the cause of unexplained
atelectasis in i 5. The patients’ medical records were reviewed, and finding of mediastinal lymphadenopathy in five tumor cases
the cause of atelectasis was established by findings from fiberoptic did not increase the yield but did increase the confidence of
bronchoscopy (36 cases), percutaneous needle biopsy (two cases), the diagnosis.
thoracotomy (two cases), thoracentesis (four cases), or clinical follow- Evaluation of the CT scans showed that air bronchograms
up(six cases). Segmental and lobar atelectasis was evenly distributed were present in 32 of the 50 patients and were absent in 18.
among all lobes of both lungs. Twenty-seven patients had collapse In the 32 with air bronchograms, an obstructing tumor was
because of an obstructing tumor (26 had bronchogenic carcinoma present in 1 1 (34%) and absent in 21 (66%). In the 1 8 without
and one had Iymphoma). Fiberoptic bronchoscopy had been per-
air bronchograms, an obstructing tumor was responsible in
formed in 26 of these 27. In 2i cases, visual findings at fiberoptic
1 6 (89%) whereas tumor was absent in only two (1 1 %).
bronchoscopy showed an endobronchial tumor mass, and biopsy of
Although the demonstration of air bronchograms on CT was
the mass was positive for malignancy. In five cases, visual findings
at fiberoptic bronchoscopy showed narrowing
of the bronchus with not considered to have diagnostic significance, the absence
occlusion but no definite endobronchial
Brushings, washings, mass. of air bronchograms on CT was thought to favor obstructing
or biopsies taken at fiberoptic bronchoscopy were positive for malig- tumor (Fig. 4). False-positive diagnoses occurred in one case
nancy in three of these cases and were equivocal in two. In these of pneumonia in which the bronchi were filled by secretions
 AJR:150, April 1988 PULMONARY ATELECTASIS 759

evident that central bronchial abnormality took precedence

over this observation.
A central hilar mass was evident by CT in 25 cases (Figs.
3 and 5). All 25 were due to obstructing tumor. Two of the
27 tumors did not show a central mass on CT; none of the
other causes of atelectasis showed a central hilar mass.
Central bronchial abnormality, consisting of either bronchial
narrowing or cutoff (Figs. 3-5), was identified in 30 cases. All
27 of the obstructing tumors were included in this group; the
CT findings in these 27 cases were confirmed by visual and/
or pathologic findings from fiberoptic bronchoscopy in 24,
visual fiberoptic bronchoscopy and pathology from percuta-
neous needle biopsy in two, and thoracotomy in one. How-
ever, there were three false-positive diagnoses-one case of
tuberculous bronchostenosis with bronchial narrowing, one
case of pneumonia with obstructing secretions, and one case
of massive pleural effusion with bronchial compression. All 20
cases with normal central bronchi on CT were correctly
diagnosed as not being caused by obstructing tumor (Figs. 1
and 2).
Either a central hilar mass or central bronchial abnormality
American Journal of Roentgenology 1988.150:757-763.

was considered to be a CT sign of atelectasis from obstructing

tumor sensitivity was 1 00% (27/27) and specificity was 87%
(20/23). The false-negative rate was 0% (0/20), and the false-
positive rate was 1 0% (3/30). Mediastinal lymphadenopathy
was evident in 1 9 cases, 1 8 of which were associated with
malignant tumors. Again, this did not increase the yield of
tumor diagnosis but did increase confidence of diagnosis in
these cases.
In 35 cases, the cause of atelectasis was reasonably certain
from the history, clinical findings, and chest radiographic
findings; however, in 1 5 (30%), the cause of atelectasis was
not apparent on the basis of these factors (Fig. 2). In these
1 5 cases, the following conditions were present: pneumonia
or other acute air-space disease in four, obstructing tumor in
three, pleural effusion in three (malignant in two and nonma-
lignant in one), bronchiectasis in three, and radiation fibrosis
in two. CT successfully detected the three missed carcinomas
(Fig. 4) and successfully excluded obstructing tumor in 1 0 of
the remaining 1 2 cases (Fig. 2), giving a correct diagnosis in
87% of the cases in which the cause of atelectasis had been
Fig. 1.-Atelectasis due to blunt chest trauma. unknown. Also, in the five patients with compressive atelec-
A, Chest radiograph shows complete left lower lobe atelectasis. Nu- tasis caused by pleural effusion, CT showed pleural tumor
merous air bronchograms (arrowheads) and patent central bronchi (arrow)
are present.
spread in two, correctly indicating the malignant nature of
B, CT scan shows a patent lower lobe bronchus (arrow) with no central effusion.
mass, excluding obstructing tumor as cause of atelectasis.

Discussion

and in one case of compressive atelectasis from massive The chest radiograph is an excellent diagnostic tool for
pleural effusion in which the lobar bronchus had collapsed showing segmental or lobar atelectasis. The exclusion of a
from extrinsic pressure. central obstructing tumor as the cause of atelectasis rests
“Mucous” bronchograms were seen in 1 2 of the 50 cases. primarily on showing that the central bronchi are patent [3]
CT showed central bronchial obstruction from tumor in nine (Fig. 1). However, in many cases the proximal lobar or seg-
cases (75%); in three cases (25%) in which obstruction was mental bronchi are not visualized adequately to show definite
due to pneumonia or abnormal mucociliary transport, CT patency or occlusion. When this is the case, other techniques
showed patent central bronchi. Although mucous broncho- are required to evaluate the bronchi. Although overpenetrated
grams suggested obstructing carcinoma (Fig. 5), the total or oblique films may improve visualization of the central
number of cases with this finding was small, and it was bronchi, conventional tomography, bronChography, or bron-
 760 WOODRING AJR:150, April1988
American Journal of Roentgenology 1988.150:757-763.

Fig. 2.-Atelectasis due to tuberculosis.

A, Chest radiograph shows increased density
in right upper lobe due to complete atelectasis
of apical segment of right upper lobe (arrow).
No definite central mass or bronchial cutoff is
seen. Differential diagnosis also includes apical
pleural thickening and apical carcinoma.
B, Baseline radiograph made 2 years earlier.
C and 0, CT scans show that apical segment
bronchus is patent (arrow in C), excluding ob-
structing tumor. Numerous dilated air broncho-
grams are present in atelectatic apical segment
(arrow in D), indicating bronchiectasis as the
cause of atelectasis. Sputum cultures subse-
quently grew Mycobacterlum tuberculosis; all
studies for bronchogenic carcinoma were nega-

choscopy have usually been recommended for this purpose it was generally less sensitive, indicating that in some cases
[3]. It is apparent from this study that CT is accurate enough the chest radiograph cannot be used to determine if a central
in patients with segmental or lobar atelectasis, that it should tumor is the cause of atelectasis.
supplant conventional tomography and bronchography en- CT has been shown to be an excellent means of visualizing
tirely, and that it may be used to determine which patients the lobar bronchi and proximal portions of the segmental
need bronchoscopy. bronchi in normal subjects [5]. CT can show the full length of
The chest radiograph can be used to document central every major bronchus that courses horizontally, including their
tumor as the cause of atelectasis in most cases in which origin and proximal portion, but bronchi that have a cephalic-
tumor is present on the basis of two major findings: evidence caudal course can only be seen in cross-section [5]. Only
of central hilar mass (S sign of Golden) and evidence of central those bronchi that run obliquely (lingular bronchus and its
bronchial narrowing or occlusion [3, 4]. Using both findings inferior and superior divisions) may be difficult to visualize in
as plain film signs, we found that the chest radiograph had some cases [5].
96% specificity for obstructing tumor and a low false-positive Several studies have addressed the ability of CT as com-
rate. Although the chest radiograph was more specific than pared with that of fiberoptic bronchoscopy to detect endo-
CT in identifying obstructing tumor as the cause of atelectasis, bronchial disease in a large number of benign and malignant
 AJR:150, April1988 PULMONARY ATELECTASIS 761
American Journal of Roentgenology 1988.150:757-763.

Fig. 3.-Atelectasis due to carcinoma. Posteroanterior (A) and lateral (B) chest radiographs
show left upper lobe atelectasis with bronchial cutoft (solid arrow) and evidence of central hilar
mass (open arrow).
C, CT scan shows central hilar mass (arrow) with bronchial Irregularity and cutoff of lingular
segmental bronchus of the left upper lobe (arrowhead), indicating obstructing tumor as the cause
of atelectasis. Higher sections revealed occlusion of remainder of left upper lobe bronchus but did
not show the hilar mass.

diseases, including cases with peripheral lung masses, pul- missed CT diagnoses came from this group [61. Naidich et al.
monary consolidation, atelectasis, hilar and mediastinal [7] studied 64 patients with endobronchial lesions at fiberoptic
masses, pleural effusion, and isolated endobronchial abnor- bronchoscopy by CT and found excellent correlation between
malities [6, 7]. Henschke et al. [6] compared both techniques the two methods in 59 (92%). In five, CT missed or underes-
in 46 patients with malignant disease and in 54 patients with timated the extent of the endobronchial lesion; most of the
benign disease. In nine of the 46 patients with malignant missed lesions were in the left bronchial tree. CT was also
disease, CT and visual findings at fiberoptic bronchoscopy unable to determine whether the abnormality was primarily
disagreed. In five, CT showed a bronchial abnormality, but endobronchial, submucosal, or extrinsic, and was not histo-
fiberoptic bronchoscopy showed no definite endobronchial logically specific [7]. Only 1 8 (28%) of the 64 had atelectasis;
mass; in four, brushings, washings, or biopsies at the site of none of the missed lesions came from this group [7]. Although
the abnormality were positive for malignancy, and in one both studies found overall excellent correlation between CT
case, malignancy was revealed by thoracotomy. In four cases, and fiberoptic bronchoscopy, there were examples of missed
CT was negative for bronchial abnormality, but fiberoptic or underestimated endobronchial lesions by CT [6,7]. Naidich
bronchoscopy revealed small endobronchial lesions. Only 17 et al. [7] recommended that CT be used as a screening
(37%) of the 46 malignant cases had atelectasis; none of the procedure only for those patients with a low index of suspicion
 762 W000RING AJR:150, April 1988

Fig. 4.-Atelectasis due to carcinoma. Fig. 5.-Squamous cell carcinoma of bronchus intermedius with complete atelectasis of right
Patient had unexplained atelectasis of anterior middle and lower lobes.
segment of right upper lobe discovered on chest A, CT scan I cm below carina shows a low-density central mass (arrow) with complete obstruction
radiograph. CT scan shows atelectasls of segment of bronchus intermedius.
without air bronchograms (arrow). Although ante- B, CT scan 1 cm lower shows numerous branching mucous bronchograms within atelectatic right
her segment bronchus could not be visualized, no lower lobe (arrowheads).
American Journal of Roentgenology 1988.150:757-763.

central hilar mass was evident. Absence of air

bronchograms and inability to identity segmental
bronchus on CT were felt to Indicate obstructing
tumor as the cause of atelectasis. Bronchoscopy
visually showed occlusion of anterior segmental
bronchus without a definite endobronchlal mass;
however, brush biopsy of segmental bronchus re-
vealed non-oat cell carcinoma.

for endobronchial disease (such as young patients with hem- similar conditions-bronchial compression from large pleural
optysis or infection); they also stated that a negative CT scan effusion, bronchial stricture, and filling of bronchi by inflam-
cannot be assumed to be definitive for patients in whom there matory secretions. However, the rate of false-positive diag-
is a strong clinical suspicion of endobronchial disease. How- noses was slightly higher (10%).
ever, if one considers the subgroup of 35 patients with CT is reliable in diagnosing obstructing tumor as a cause
atelectasis in these two studies, bronchial obstruction from of atelectasis with an acceptable false-positive rate. In certain
tumor was not missed in any case by CT [6, 7]. This was patients, such as those who develop acute postoperative
confirmed by Naidich et al. [1 ] in a separate report of 36 atelectasis, those with clinically documented pneumonia, or
patients with atelectasis caused by obstructing tumor and in those with strong plain film evidence of an obstructing tumor,
the 27 cases in this report. Therefore, it appears that when determination of the cause of atelectasis is usually not a
an endobronchial lesion is large enough to result in postob- diagnostic problem. However, 1 5 (30%) of the patients in this
structive atelectasis, CT should be reliable in detecting its series had initial evidence of atelectasis with no certainty
presence (98 of 98 reported cases). about the cause. By showing patent bronchi and the absence
Evidence of a central hilar mass or of bronchial narrowing of a central mass on CT, obstructing tumor can be reliably
or occlusion have been shown to be the major CT findings, excluded as the cause of atelectasis [1 9, 1 0, 1 2]. Care must
,

indicating a central tumor as the cause of atelectasis [1 4, , be taken to show the full length of the lobar or segmental
8-1 2]. Naidich and coworkers [1 9, 1 0, 1 2] have shown that
, bronchi. In bronchi that course horizontally this is not a
central bronchial abnormalities are the most important CT problem; however, in bronchi that are oriented cephalad-
sign of obstructing tumor; this was true in our series as well. caudad or obliquely, this can be difficult in some cases [12].
Combining CT evidence of central hilar mass or central bron- Careful attention should be paid to the technique used in the
chial narrowing or occlusion resulted in a correct diagnosis in CT examination. Evaluation of the lobar and segmental bron-
all cases of obstructing tumor (Figs. 3-5). chi by contiguous 5-mm collimated scans or by overlapping
Naidich and coworkers [1 9, 1 0] addressed , the false- scans may be crucial in preventing erroneous diagnoses of
positive rate of CT diagnosis in pulmonary atelectasis and obstructing tumor beCause of either partial-volume averaging
found it to be low (5%). In their experience, false-positive artifact or missed subtle bronchial abnormalities [1 0, 1 2]. In
diagnoses of tumor by CT occurred as a result of bronchial most cases in this study, contiguous 1 0-mm collimated scans
narrowing or occlusion in cases of bronchial stricture, mucous were adequate; however, when there was any question as to
plugging, and bronchial compression from tension pneumo- whether the bronchi were normal in the atelectatic lobe or
thorax. In this study, false-positive diagnoses occurred from segment, the thinner scans were used. Because of the ex-
 AJR:150, April 1988 PULMONARY ATELECTASIS 763

cellent correlation between CT and fiberoptic bronchoscopy the further staging of disease in patients with obstructing
and pathologic findings, it is unlikely that any significant tumor [10].
lesions were missed by this technique in this group of patients
with atelectasis. Also, contrast enhancement during the study REFERENCES
may be helpful in distinguishing central hilar masses from
1 . Naidich DP, McCauley Dl, Khouri NF, Leitman BS, Hulnick DH, Siegelman
vessels and in showing tumor within collapsed parenchyma
Ss. Computed tomography oflobar collapse. 1 . Endobronchial obstruction.
[10]. J Comput Assist Tomogr 1983:5:745-757
In addition, specific diagnoses such as bronchiectasis (Fig. 2. Proto AV, Tocino I. Radiographic manifestations of lobar collapse. Semin
2) or compressive atelectasis from pleural effusion may be Roentgenol 1980:15:117-173
3. Felson B. Chest roentgenology. Philadelphia: Saunders, 1973
made by CT. In this study, CT clarified most of these cases.
4. Reinig JW, Ross P. Computed tomography appearance of Goldens S
As Naidich et al. [1 0] have indicated, CT is also helpful in the sign. J Comput Tomogr 1984:8:219-223
evaluation of compressive atelectasis caused by pleural ef- 5. Naidich DP, Terry PB, Stitick FP, Siegelman SS. Computed tomography
fusion. In 22 of 23 cases [1 0], CT suggested the malignant of the bronchi. 1 . Normal anatomy. J Comput Assist Tomogr 1980:4:
nature of pleural effusion by showing nodular pleural tumor 746-753
6. Henschke Cl, Davis SD, Auh Y, et al. Detection of bronchial abnormalities:
deposits and loculated collections of pleural fluid not sus-
comparison of CT and bronchoscopy. J Comput Assist Tomogr
pected on plain films. In this study, CT correctly suggested 1987:11:432-435
the malignant cause of pleural fluid in both cases. 7. Naidich DP, Lee J-J, Garay SM, McCauley DI, Aranda CP, Boyd AD.
Therefore, the indications for CT in the evaluation of seg- Comparison of CT and fiberoptic bronchoscopy in the evaluation of bron-
mental and lobar atelectasis are as follows: (1) CT should be chial disease. AJR 1987:1 48: 1-7
8. Tobler J, Levitt AG, Glazer HS, Moran J, Crouch E, Evens AG. Differentia-
performed in any case in which the cause of atelectasis is tion of proximai bronchogenic carcinoma from postobstructive lobar ccl-
uncertain. (2) Endobronchial obstruction may be differentiated lapse by magnetic resonance imaging: comparison with computed tomog-
American Journal of Roentgenology 1988.150:757-763.

from the other causes of atelectasis in most cases, negating raphy. Invest Radiol 1987:22:538-543
the need for bronchoscopy in those patients with normal 9. Naidich DP, McCauley Dl, Khoun NF, Leitman BS, Hulnick DH, Siegelman
SS. Computed tomography of lobar collapse. 2. Collapse in the absence
bronchi and indicating the need for bronchoscopy with biopsy
of endobronchial obstruction. J Comput Assist Tomogr 1983:7 :758-767
for histologic diagnosis in those with abnormal bronchi [9, 1 0. Naidich DP, Ettinger N, Leitman BS, McCauley Dl. CT of lobar collapse.
12]. (3) CT should be performed as part of the evaluation of Semin Roentgenol 1984:19:222-235
atelectasis associated with pleural effusion. Endobronchial 1 1 . Raasch BN, Heitzman ER, Carsky EW, Lane EJ, Berlow ME, Witwer G. A
obstruction as a cause in these cases may be assessed to computed tomographic study of bronchopulmonary collapse. Radio-
Graphics 1984:4:195-232
determine whether bronchoscopy is needed; the malignant 12. Naidich DP, Stitick FP, Khouri NF, Terry PB, Siegelman SS. Computed
nature of pleural effusion may be indicated, focusing the tomography of the bronchi. 2. Pathology. J Comput Assist Tomogr
workup on the pleural space [9]. CT may also be helpful in 1980;4:754-762