AJR 2000; 175:477-483
© American Roentgen Ray Society
Differentiating Normal from Abnormal Inferior Thoracic Paravertebral Soft Tissues on Chest Radiography in Children
Lane F. Donnelly1,2,
Donald P. Frush1,
Jing-Yuan Zheng1 and
George S. Bisset, III1
1
Division of Pediatric Radiology, Department of Radiology, Duke University
Medical Center, Durham, NC 27710.
2
Present address: Department of Radiology, Children's Hospital Medical Center
and the University of Cincinnati, 3333 Burnet Ave., Cincinnati, OH
45229-3039.
Received October 22, 1999;
accepted after revision January 6, 2000.
Address correspondence to L. F. Donnelly.
Abstract
OBJECTIVE. The purposes of this investigation were to define the
normal appearances, define factors that have the potential to influence
appearance, and establish defined criteria to differentiate normal from
abnormal appearances of posteroinferior paravertebral soft tissues on chest
radiography in children.
SUBJECTS AND METHODS. Paravertebral soft tissues were evaluated on
frontal chest radiographs in 23 children with documented abnormalities and 275
children without abnormalities in the region. The frequency of visualization,
course, width, and factors (patient positioning, age, and sex) potentially
influencing the appearance of paravertebral soft tissues were determined.
Inferolateral course and width greater than that of the adjacent pedicle were
evaluated as criteria for abnormality.
RESULTS. Only 28% of the children without abnormalities had
paravertebral soft tissues visualized, and the frequency of visualization
directly increased with age (p = 0.001). For identification of
abnormal cases on the left side, width greater than the adjacent pedicle had a
sensitivity of 100% and a specificity of 98%, and inferolateral course had a
sensitivity of 86% and a specificity of 95%. Visualization on the right side
(n = 5) was always abnormal. Six normal cases had a width greater
than that of the adjacent pedicle on the left side on initial radiographs
obtained with supine positioning and met normal criteria on repeated
radiographs with upright positioning.
CONCLUSION. Width greater than the adjacent pedicle is the best
radiographic criterion for differentiation of abnormal from normal left-sided
paravertebral soft tissues, particularly on radiographs obtained with upright
positioning. Identifiable right-sided paravertebral soft tissue is always
abnormal. These criteria are useful aids in determining the need for
additional imaging, such as CT.
Introduction
Accurate evaluation of the paravertebral soft tissues of the inferior
thorax is of particular importance when interpreting radiographs of children
because this region can be affected by a number of pathologic processes, most
commonly neuroblastoma
[1,2,3].
When a pathologic process involves the region, the paravertebral soft tissues
become widened and assume a convex appearance on frontal radiographs of the
chest [3]. When paraspinal
masses are large, differentiating the abnormality seen on radiography from the
normal appearance of the paraspinal soft tissues is not difficult. However,
radiographic differentiation of abnormal from normal can be difficult when the
abnormality is subtle. We have encountered multiple cases in which the left
paravertebral soft tissues appear only mildly widened or are slightly convex
with an inferolateral course. In our experience, further examination with CT
has revealed abnormalities in some cases (Fig.
1A,1B)
and no abnormality in others (Fig.
2A,2B).
Whether children with these subtle findings need to undergo CT or return for
follow-up chest radiography is a clinical issue. The potential of missing a
significant abnormality is weighed against performing unnecessary CT or MR
imaging examinations. The establishment of criteria to aid in differentiating
normal from abnormal paravertebral soft tissues would facilitate appropriate
imaging evaluation. The spectrum of normal appearances of the paravertebral
soft tissues, the changes with age, and the effects of upright versus supine
positioning on the appearance of the paravertebral soft tissues have not been
investigated to our knowledge. The purposes of this investigation were to
define the normal appearances of the posteroinferior paravertebral soft
tissues on chest radiography in children, define factors that influence the
appearance, and establish criteria that differentiate normal from abnormal
appearance.
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —2-year-old girl with neuroblastoma who presented with malaise and
abdominal pain. Radiograph shows bilateral paraspinal soft tissues (large
arrows), both of which are oriented inferolaterally and greater in width
than width of adjacent pedicles (small arrows).
|
|
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —2-year-old girl with neuroblastoma who presented with malaise and
abdominal pain. CT scan, obtained same day as A, at most inferior level
of posterior lung shows large heterogeneous mass, representing neuroblastoma,
engulfing aorta and celiac artery. Mass contributes to widening of paraspinal
soft tissues.
|
|
View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —7-month-old girl who presented with malaise and weight loss and was
falsely thought to have paraspinal mass on radiography. Radiograph shows left
paraspinal soft tissues (arrow) to be oriented inferolaterally. Width
of left paravertebral soft tissues is slightly less than that of adjacent
pedicle. No right paravertebral soft tissues are visualized. Note right aortic
arch.
|
|
View larger version (96K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —7-month-old girl who presented with malaise and weight loss and was
falsely thought to have paraspinal mass on radiography. CT scan, obtained
within 1 hr of chest radiograph, shows normal findings. There is no paraspinal
mass or other explanation for appearance of left paraspinal soft tissues.
|
|
Subjects and Methods
The inferoposterior paravertebral soft tissues were evaluated on frontal
chest radiographs in 23 children with documented pathology and in 275 children
without pathology.
The 275 children without abnormality on chest radiographs of the
paravertebral soft tissues were prospectively enrolled in the study. The mean
age of the children was 7 years and the age range was 1 day to 18 years. A
data sheet was filled out by one of three potential reviewers when the
radiograph was clinically interpreted.
The chest radiographs of these 275 children were obtained for various
reasons, most commonly for the exclusion of pneumonia or pneumothorax or for
the evaluation of wheezing. There were certain exclusion criteria. Because an
abnormal appearance of the paravertebral soft tissues has been described on
radiographs obtained with rotated positioning
[1,
4], children whose radiographs
were obtained with the patient rotated were not enrolled in the study.
Radiographs were excluded when there was rotation to the extent that there was
a ratio greater than 2:1 comparing the left and right distances between the
clavicular heads and the transverse process of the isolevel vertebral body.
Other reasons for exclusion from the study group included the presence of
scoliosis [1,
3], previous thoracic surgery,
or the presence of adjacent lung opacification obscuring visualization of the
paravertebral soft tissues. In all cases, the age and sex of the patient,
technique, and patient positioning at the time of the chest radiograph were
noted. To limit the number of technical variables, all chest radiographs in
the study were obtained with either upright positioning and posteroanterior
technique or supine positioning and anteroposterior technique.
Because all the children in this study had clinical indications for the
chest radiographs to be obtained, the patient population did not represent the
optimal normal population. Patients with low-acuity medical problems, such as
patients with chest radiographs obtained at outpatient primary care centers
and the emergency room, were targeted for inclusion in the study. Criteria for
normalcy included lack of clinical, radiographic, or other imaging findings
that suggested the possibility of a paraspinal process.
The chest radiographs of the 23 children with documented abnormalities were
compiled through a combination of retrospective collection of previous cases
from the authors' personal and departmental teaching files and the prospective
addition of cases with abnormal findings encountered during the study. The
causes of the abnormal paraspinal widening included neural crest tumors
(n = 5) (ganglioneuroma [Fig.
3A,3B],
neuroblastoma [Figs.
1A,1B
and
4A,4B]),
lymphoma (n = 1), extramedullary hematopoiesis (n = 1),
duplication cyst (n = 1), neurenteric cyst (n = 1), hiatal
hernia (n = 1), diskitis (n = 4), hematoma after trauma with
or without aortic dissection (n = 2) (Fig.
5A,5B),
abnormal course of hepatic vein confluence to right atrium caused by azygous
continuation of the inferior vena cava (n = 1) (Fig.
6A,6B),
medial pleural fluid (n = 5), and chronic inflammatory pleural
thickening (n = 1). The paraspinal involvement by the cases of
neuroblastoma, lymphoma, extramedullary hematopoiesis, duplication cyst, and
chronic inflammatory pleural thickening were documented with CT or MR imaging,
and the diagnoses were confirmed at surgery. Other imaging procedures, without
pathologic confirmation, were used to document the cases of azygous
continuation (MR imaging), hiatal hernia (fluoroscopic barium study), diskitis
(MR imaging), aortic dissection with hematoma (CT and arteriography), and
medial pleural fluid (radiography with decubitous positioning, sequential
radiography, or sonography).
View larger version (118K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —14-year-old girl with left paraspinal ganglioneuroma. Radiograph
shows inferolateral course of left paraspinal soft tissues. Width of left
paravertebral soft tissues (arrows) is greater than width of adjacent
bony pedicle (arrowheads). No right paraspinal soft tissues are
visualized.
|
|
View larger version (138K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —2-year-old girl with right paraspinal neuroblastoma. Radiograph
shows presence of right paravertebral soft tissues (arrows).
Paravertebral soft tissues are wider than adjacent pedicle
(arrowheads). No left paravertebral soft tissues are present.
|
|
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —15-year-old girl with mediastinal hematoma after high-speed motor
vehicle collision. Radiograph shows widening of left paraspinal soft tissues
(arrows), which are vertically oriented and more than twice width of
adjacent pedicle. No right paraspinal soft tissues are visualized.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —15-year-old girl with mediastinal hematoma after high-speed motor
vehicle collision. CT scan shows abnormal paravertebral soft tissues
(arrow) elevating descending aorta away from adjacent vertebral body
and causing appearance seen on radiography. Aortogram (not shown) showed
normal findings. Abnormal soft tissue was presumed to be hematoma related to
venous bleeding.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A. —15-year-old girl with abnormal course of hepatic vein confluence to
right atrium caused by azygous continuation of inferior vena cava. Radiograph
shows presence of right paraspinal soft tissues (arrows), which are
oriented inferolaterally and are wider than adjacent pedicle.
|
|
View larger version (161K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B. —15-year-old girl with abnormal course of hepatic vein confluence to
right atrium caused by azygous continuation of inferior vena cava. Axial
T1-weighted MR image (TR/TE, 869/16) shows confluence of hepatic veins (H)
forming interface with aerated lung. Azygous continuation of inferior vena
cava is evident by enlarged azygous vein (V). A = aorta.
|
|
On all frontal radiographs, the presence of identifiable left- and
right-sided paravertebral soft tissues was evaluated in the inferoposterior
mediastinum, just above the hemidiaphragms. When identified, the
superior-to-inferior course and width of the paravertebral soft tissues were
characterized. The superior-to-inferior course was noted as oriented
inferomedially, vertically, or inferolaterally (Figs.
1A,1B,2A,2B,3A,3B,4A,4B,5A,5B).
The width was noted to be greater or less than that of the bony pedicle of the
adjacent vertebral body (Fig.
3A,3B).
The level at which the width was measured was that at which the paravertebral
soft tissues were the widest. On the left side, when paravertebral soft
tissues were visualized, whether a separate distinct descending aorta could be
identified at the level of the cardiophenic angle was also noted.
For the normal cases, the frequency at which paravertebral soft tissues
were identified was recorded. Differences in frequency, course, and width were
compared with patient age, patient sex, and the two radiographic techniques.
Statistically significant differences between subgroups were evaluated with a
Fisher's exact test.
Differences in course and width were evaluated as criteria to differentiate
the normal from the abnormal cases. The sensitivity and specificity for width
greater than that of the adjacent pedicle, inferolateral course, and the
combination of both were calculated. The effect of radiographic technique
(posteroanterior and upright versus anteroposterior and supine) on the
sensitivity and specificity of these criteria was also evaluated.
Results
For the 23 abnormal cases, paravertebral soft tissues were identified on
the left side in 21 patients and on the right side in five patients. In three
cases (two neuroblastomas, one neurenteric cyst), paravertebral soft tissues
were identified bilaterally (Fig.
1A,1B).
Of the 21 cases in which paravertebral soft tissues were identified on the
left side, the course was inferolateral in 18, vertical in three, and
inferomedial in none. In all 21 left-sided cases, the width of the
paravertebral soft tissues was greater than that of the adjacent pedicle. Of
the five right-sided cases, all were categorized by an inferolateral course
and width greater than that of the adjacent pedicle (Figs.
1A,1B,
4A,4B,
and
6A,6B).
For the chest radiographs in which no abnormality of the paravertebral soft
tissues was identified or suspected clinically, right-sided posteroinferior
paravertebral soft tissues were never visualized. Left-sided posteroinferior
paravertebral soft tissues were identified in 28% (76/275) of cases. Of those
76 cases, the course was inferomedial in 44, vertical in 18, and inferolateral
in 14. The width was greater than that of the adjacent pedicle in six and less
than that of the adjacent pedicle in 70. In only 13 of the 76 cases in which
left paravertebral soft issues were identified, a separate distinct descending
aorta was identified in the inferior paraspinal region. Therefore, the
descending aorta could be identified in the region of the cardiophrenic angle
in only 5% (13/275) of the cases (Figs.
7 and
8A,8B).
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8A. —Change in appearance of paraspinal soft tissue with different
radiographic technique and patient positioning in 14-year-old girl. Radiograph
obtained with anteroposterior technique and supine positioning shows left
paraspinal soft tissues (arrows) as prominent with inferolateral
orientation and width greater than adjacent pedicle.
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8B. —Change in appearance of paraspinal soft tissue with different
radiographic technique and patient positioning in 14-year-old girl. Radiograph
obtained with posteroanterior technique and upright positioning shows
resolution of prominence of left paraspinal soft tissues (arrows)
with inferomedial course and minimal width, which is less than that of
adjacent pedicle.
|
|
For the 275 normal cases, the relationship between patient age and
frequency at which left paraspinal soft tissues were identified was
statistically significant (p = 0.001). Left-sided paravertebral soft
tissues were identified in 9% (6/65) of patients younger than 1 year old. In
contrast, for patients 11-18 years old, the frequency was 42% (37/88). For
patients 1-5 years old and 6-10 years old, the frequency was 26% (18/70) and
29% (15/52), respectively. The relationship between patient sex and frequency
of visualization was not statistically significant (p > 0.5). Left
paravertebral soft tissues were seen in 24% (28/117) of girls and 30% (48/158)
of boys. Likewise, there was no statistically significant relationship between
radiographic technique and frequency of visualization (p > 0.5).
Left paravertebral soft tissues were seen in 25% (39/153) of radiographs with
posteroanterior technique and upright patient positioning and 30% (37/122) of
radiographs with anteroposterior technique and supine positioning.
When left paravertebral soft tissues were identified, radiographic
technique and patient positioning influenced the visualized width. For
posteroanterior radiographs with upright positioning, the width of the
paravertebral soft tissues was never greater than that of the adjacent pedicle
(0/153). For anteroposterior radiographs with supine positioning, the
paravertebral soft tissues were wider than the adjacent pedicles in six of
122. The width of the paravertebral soft tissues was less than the adjacent
pedicles in all six of these patients on repeated radiographs obtained with
posteroanterior technique and upright positioning
(Fig. 7). Five of these six
patients were 11-18 years old. Inferolateral course was seen in seven of the
153 radiographs obtained in the posteroanterior group and in seven of the 122
radiographs obtained in the anteroposterior group.
We noted the sensitivities and specificities of radiographic parameters in
differentiating abnormal from normal. On the left side, when inferolateral
course was used as the criterion to differentiate between the 275 normal cases
and the 21 cases with abnormal left paravertebral soft tissues, there were 18
true-positive abnormalities, three false-negative abnormalities, 14
false-positive abnormalities, and 261 true-negative abnormalities. These
results correspond with a sensitivity of 86% and a specificity of 95%. When
width greater than that of the adjacent pedicle was used as the criterion for
abnormality, there were 21 true-positives, zero false-negatives, six
false-positives, and 269 true-negatives. This corresponds with a sensitivity
of 100% and a specificity of 98%. When both criteria—width greater than
that of the adjacent pedicle and inferolateral course—are used, the
sensitivity is 86% and the specificity is 98%. All five cases in which
right-sided posteroinferior paravertebral soft tissues were visualized were
abnormal. Therefore, when visualization of right paravertebral soft tissues
was used as the criterion for abnormality, the sensitivity was 22% and the
specificity was 100%.
Because radiographic technique and patient positioning influenced apparent
course and width, the specificities are different only when radiographs with
posteroanterior technique are considered. For width greater than that of the
adjacent pedicle, there were zero false-positives and 153 true-negatives with
a specificity of 100%. For inferolateral course, we noted seven
false-positives and 146 true-negatives with a specificity of 95%. Because the
abnormal group was not altered, true-positives, false-negatives, and
sensitivity were unchanged.
Discussion
The radiographic appearance of the paravertebral soft tissues was first
detailed in a series of articles in the early 1940s
[5,6,7,8].
The paraspinal soft tissues are seen on frontal radiographs of the chest as
interfaces between soft tissue and air that parallel the borders of the mid
and lower thoracic vertebrae (paraspinal lines)
[5,6,7,8,9,10,11,12].
That the left paraspinal line is much more often identified than the right is
well documented [7,
10,
13]. This is related to
several factors. The left paraspinal position of the descending aorta deviates
the left-sided paraspinal soft tissues more laterally than those on the right
side. This brings the interface between aerated lung and paraspinal soft
tissue into a sagittal plane, rendering it visible as a line on frontal
radiographs [7,
13]. The right paraspinal soft
tissues gradually slope posteriorly, producing an oblique interface that is
not well visualized on frontal radiographs
[7,
13]. Also contributing to the
more frequent visualization of the left paraspinal line is the more abundant
mediastinal fat on the left side, surrounding the descending aorta
[10].
In adults, the typical appearance of the paraspinal soft tissues includes
visualization of the left paraspinal line, visualization of a separate
discrete shadow of the descending aorta
[10], and nonvisualization of
the right paraspinal line. This study has shown that this is not always the
case in children. In our series, the left paraspinal line was identified in
the inferior hemithorax in only 28% of children. In addition, a separate
distinct descending aorta was identified extending into the cardiophrenic
sulcus in only 5% (13/275) of patients. The frequency at which the left-sided
paraspinal soft tissues were visualized was related to age. Because of the
extremely low frequency at which the paraspinal soft tissues are seen in
younger children, paraspinal soft tissues that meet abnormal criteria should
be considered of extreme importance in young children. In adults, it has been
noted that the left-sided paravertebral soft tissues are seen more frequently
and often appear more prominent as a normal phenomenon of aging
[10,
13]. This is thought to be
related to a number of factors including increasing tortuosity of the aorta
with increasing protrusion into the left hemithorax, increased deposition of
mediastinal fat, and osteophyte formation
[10,
13]. The absence of such
factors may contribute to the low frequency of visualization of the left-sided
paraspinal soft tissues in children. In both adults and children, the
descending aorta moves inferomedially and is usually in a prevertebral rather
than left paravertebral position when it exits the thorax via the aortic
hiatus [10]. Because of this,
the descending aorta can lose its interface-forming border with the adjacent
lung [10]. The lack of
tortuosity or ectasia seen in children, compared with that seen in adults, may
also explain the decreased frequency with which the descending aorta was
identified in our series. In our study, right-sided paraspinal soft tissues
were never visualized in normal cases.
Although there was initial debate about the true cause of the paraspinal
lines
[5,6,7,
10], the diagnostic importance
of increased width and convexity of the paraspinal lines was recognized early
[7,
8]. In children, there are
multiple causes of paraspinal widening
[1,
2]. In 1967, the importance of
widening of the paraspinal soft tissues in the lower thorax and upper lumbar
regions was described in the diagnosis of neuroblastoma
[3]. Neuroblastoma remains the
most common cause of such paraspinal widening in children
[1,
2]. Other neoplastic causes of
paraspinal widening in children include lymphoma
[1,
2], metastatic lymphadenopathy
including that caused by Wilms' tumor
[14,
15], and ganglioneuroma
[1,
2,
12,
13]. Inflammatory causes
include diskitis [1,
2,
12], vertebral osteomyelitis
[1,
2,
12,
16], lymphadenopathy resulting
from causes such as tuberculosis
[12], and abscess formation
from rupture of the esophagus
[12,
13]. Vascular causes include
hematoma from aortic dissection or spinal injury, aortic aneurysm, and
enlargement of the azygous-hemiazygous system
[1,
4,
12,
13]. Other described causes of
widening of the paraspinal line include extramedullary hematopoiesis
[2,
12,
13], benign lesions such as
extralobar sequestration and duplication cysts
[1], hiatus hernia
[1], and medial pleural fluid
[2,
10,
17].
Pseudotumor or factitious widening of the paraspinal line has been
described as a consequence of oblique positioning
[3,
4,
15]; therefore, radiographs
obtained with obliquity were not included in our study. In addition, the
paraspinal soft tissues can appear prominent in children with scoliosis
[1,2,3];
thus, patients with scoliosis were also not included in our study.
Because multiple clinically important diagnoses can be identified by
changes seen in the paraspinal soft tissues on frontal chest radiographs,
accurate interpretation is important. In our study, width greater than that of
the adjacent pedicle was the best radiographic criterion for differentiating
normal from abnormal cases, with a sensitivity of 100% and a specificity of
98%. Inferolateral course was also useful in excluding abnormalities, with a
sensitivity of 86% and a specificity of 95%. The presence of identifiable
posteroinferior paraspinal tissue on the right side was always abnormal. These
criteria are useful in differentiating those cases in which subtle
abnormalities distorting the paraspinal line are present from normal
variations in the appearance of the paravertebral soft tissues. The use of
these criteria should help decrease the number of CT examinations performed
for such reasons.
Supine positioning and anteroposterior technique affected the appearance of
the paravertebral soft tissues. There were six cases identified in which, on
radiographs with supine positioning, the width of the paraspinal soft tissues
was greater than that of the adjacent pedicle. On repeated radiographs with
the upright technique, the paraspinal stripe was less than that of the
adjacent pedicle. When the patient is moved from upright to supine, the change
in the axis of gravity probably exerts force on the posterior mediastinal
tissues, increasing the left-to-right axial diameter. When radiographs
obtained with the patient supine were excluded, the specificity of our width
criteria increased to 100%. Therefore, if abnormal left-sided paravertebral
soft tissues are suspected on a radiograph obtained with supine positioning
and anteroposterior technique, a second radiograph with upright positioning
and posteroanterior technique may be the next examination of choice.
This study has limitations. First, all 275 children who were considered to
have a normal appearance had clinical indications for radiography and,
therefore, did not represent the optimal normal population. In addition, there
was no gold standard such as CT or autopsy as proof of normalcy. Performing CT
or follow-up radiography for the purpose of documenting normalcy for research
purposes is unethical in healthy children. Therefore, other criteria were used
for normalcy such as a lack of imaging or clinical findings suggestive of a
paraspinal process and absence of abnormality on clinical follow-up. Another
limitation is the relatively small sample size of abnormal cases. Finally,
because only one reviewer evaluated the images, this study does not address
issues such as inter- or intraobserver reliability of the investigated
radiographic findings. This should be evaluated in future study.
In conclusion, the criteria of width greater than that of the adjacent
pedicle and inferolateral course reliably identify and exclude abnormalities
of the left paraspinal soft tissues. Any paraspinal soft tissue seen in the
right posteroinferior soft tissues is abnormal. These criteria should help in
deciding whether further imaging is needed when questionable abnormalities are
encountered on frontal radiographs of the chest.
References
-
Eklof O, Galatius-Jensen F, Damgaard-Pedersen K. Malignant versus
benign paravertebral widening in children. Pediatr
Radiol 1981;11:193
-201[Medline]
-
Gaisie G, Oh KS. Paraspinal interfaces in the lower thoracic area
in children: evaluation by CT. Radiology
1983;149:133
-135[Abstract/Free Full Text]
-
Eklof O, Gooding CA. Paravertebral widening in cases of
neuroblastoma. Br J Radiol
1967;40:358
-365[Medline]
-
Taybi H. Pseudoneoplastic masses (pseudotumours) in children. 1.
Pseudotumours of face, neck, and thorax. Med Radiogr
Photogr 1978;54:2
-20
-
Lackman E. A comparison of the posterior boundaries of the lungs
and pleura as demonstrated on the cadaver and on the roentgenogram of the
living. Anat Rec
1942;83:521
-542
-
Garland LH. The postero-mediastinal pleural line.
Radiology
1943;41:29
-37
-
Brailsford JF. The radiographic postero-medial border of the lung,
or the linear thoracic paraspinal shadow. Radiology
1943;41:34
-37
-
Billings L. Retrocardiac pulmo-pleural demarcation lines and their
diagnostic significance. Acta Radiol
1946;27:257
-261
-
Berne AS, Gerle RD, Mitchell GE. The mediastinum: normal roentgen
anatomy and radiologic techniques. Semin Roentgenol
1969;4:3
-21
-
Genereux GP. The posterior pleural reflections.
AJR
1983;141:141
-149[Abstract/Free Full Text]
-
Gupta SK, Mohan V. The thoracic paraspinal line: further
significance. Clin Radiol
1979;30:329
-335[Medline]
-
Dalton CJ, Schwartz SS. Evaluation of the paraspinal line in
roentgen examination of the thorax. Radiology
1956;66:195
-200
-
Efremidis SC, Dan SJ, Cohen BA, Mitty HA, Rabinowitz JG. Displaced
paraspinal line: role of CT and lymphography. AJR
1981;136:505
-509[Abstract/Free Full Text]
-
Eklof O, Sandstedt B, Soderlung S, Wikstad I, Ahstrom L.
Paravertebral widening in a case of recurrent Wilms tumour. Pediatr
Radiol 1982;12:255
-257[Medline]
-
Siegel MJ, McAlister WH. Unusual intrathoracic complications in
Wilms tumor. AJR
1980;134:1231
-1234[Abstract]
-
Merriman TE, Taylor RG, Nattrass GR. Vertebral osteomyelitis in an
infant presenting with a posterior mediastinal mass. Pediatr Surg
Int 1997;12:541
-543[Medline]
-
Trackler RT, Brinker RA. Widening of the left paravertebral pleural
line on supine chest roentgenograms in free pleural effusions.
AJR 1966;96
: 1027-1033[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
