Original Contribution

Reappraisal of radiographic signs of pneumoperitoneum at emergency department

Yu-Hui Chiu MD^a,b, Jen-Dar Chen MD^b,c,?, Chui-Mei Tiu MD^b,c, Yi-Hong Chou MD^b,c, David Hung-Tsang Yen MD, PhD^a,b, Chun-I Huang MD^a,b, Cheng-Yen Chang MD^b,c

^aDepartment of Emergency Medicine, Taipei Veterans General Hospital, Taiwan, ROC

^bNational Yang-Ming University School of Medicine, Taiwan

^cDepartment of Radiology, Taipei Veterans General Hospital, Taipei, Taiwan, ROC

Received 4 January 2008; revised 14 February 2008; accepted 1 March 2008

Abstract

Purpose: This study aimed to evaluate the sensitivities of the reported Free air signs on supine chest and abdominal radiographs of hollow organ perforation. We also verified the value of supine radiographic images as compared with erect chest and decubitus Abdominal radiographs in detection of pneumoperitoneum.

Methods: Two hundred fifty cases with surgically proven hollow organ perforation were included. Five hundred twenty-seven radiographs were retrospectively reviewed on the picture archiving and communication system. Medical charts were reviewed for operative findings of upper gastrointestinal tract, small bowel, or colon perforations. The variable free air signs on both supine abdominal radiographs (KUB) and supine Chest radiographs were evaluated and determined by consensus without knowledge of initial radiographic reports or final diagnosis. Erect CXR and left decubitus abdominal radiographs were evaluated for subphrenic free air or air over nondependent part of the right abdomen.

Result: Upper gastrointestinal tract perforation was proven in 91.2%; small bowel perforation, in 6.8%; and colon perforation, in 2.0%. The positive rate of free air was 80.4% on supine KUB, 78.7% on supine CXR, 85.1% on erect CXR, and 98.0% on left decubitus abdominal radiograph. Anterior superior oval sign was the most common radiographic sign on supine KUB (44.0%) and supine CXR (34.0%). Other free air signs ranged from 0% to 30.4%.

Conclusions: Most free air signs on supine radiographs are located over the right upper abdomen. Familiarity with free air signs on supine radiographs is very important to emergency physicians and radiologists for detection of hollow organ perforation.

(C) 2009

Introduction

* Corresponding author. Department of Radiology, Taipei Veterans General Hospital, Taipei, 112, Taiwan, ROC. Tel.: +886 2 28771180.

E-mail address: [email protected] (J.-D. Chen).

Hollow organ perforation is always life threatening and usually needs the urgent surgical intervention. Detection of the intraperitoneal free gas is valuable in the diagnosis of hollow organ perforation. It is really a challenge to both emergency physicians and radiologists to detect free air on

0735-6757/$ - see front matter (C) 2009 doi:10.1016/j.ajem.2008.03.004

Fig. 1 A 1-year-old female subject with cecal perforation due to chronic inflammation. The KUB shows Rigler sign (broken arrows), falciform ligament sign (straight arrows), and cupola sign (arrowheads).

supine radiographs. Because computed tomography is the most sensitive examination to visualize the extraluminal air and plain supine radiographs cannot always detect the presence and extent of an abnormal lucency, the value of the plain supine radiographs to diagnose intraperitoneal free air is usually ignored by the emergency physicians in modern times [1-4]. However, the Plain radiography is still the fastest and most efficient and cheapest imaging examination for detecting intraperitoneal free air at emergency department (ED). On plain radiographs, very small amounts of extraluminal free air still can be visualized [2,5-7]. This study aimed to evaluate the sensitivities of the reported free air signs on supine chest and abdominal radiographs of the patients with hollow organ perforation. We also verified the value of supine radiographic images as compared with erect chest radiographs and decubitus abdominal radiographs in detection of pneumoperitoneum.

Methods

Two hundred seventy-one cases with any sex, any age, and surgically proven hollow organ perforation from December 2000 to August 2007 were included in this study by electronic search of the medical registry database of our institute. After excluding the cases without available images and those with poor Image quality or intervals between radiographs and operation longer than 3 days, 250 cases were used for our study. Five hundred twenty-seven radiographs of these cases were retrospectively reviewed on the picture archiving and communication system for detection of free air, including 184

supine abdominal radiographs (KUB), 150 supine chest radiographs (CXR), 94 erect CXR, and 99 left decubitus abdominal radiographs. Medical charts were reviewed for the operative findings of perforation sites, which were categor- ized into 3 groups of upper gastrointestinal tract (UGI), small bowel, and colon perforations. The underlying etiologies were also recorded.

The variable free air signs on the supine radiographs reported in the literature were evaluated for every case, including Rigler sign, hyperlucent liver sign, falciform ligament sign, inverted V sign, urachus sign, anterior superior oval sign, fissure for ligament teres sign, doge cap sign, football sign, cupola sign, hepatic edge sign, triangle sign, and dolphin sign. Besides, the unnamed focal radiolucency and left-sided anterior superior oval sign on the supine radiographs (both supine KUB and supine CXR) and subphrenic radiolucency on the supine CXR, despite being not well documented in previous reports, were also evaluated. These free air signs were categorized into 4 groups: bowel- related signs, right-upper-quadrant (RUQ) signs, peritoneal ligament-related signs, and other signs. All of the above signs were searched for and decided by consensus of 2 experienced radiologists (J.D.C. and C.M.T.) who did not know the initial radiographic reports or final diagnosis.

• 1. Definitions of radiographic signs of pneumoperitoneum
     1. Bowel-related signs
        1. Rigler sign. In the Rigler sign, also known as the bas-relief sign or the double-wall sign, both sides of the bowel wall can be visualized on a supine KUB (Fig. 1) [5,8-11].
        2. Triangle sign. Free intraperitoneal air accumulat- ing among 3 adjoining bowel loops or 2 bowel loops and the

Fig. 2 A 28-year-old female subject with perforated Duodenal ulcer. The KUB shows triangle sign (broken arrows) and fissure for ligament teres sign (straight arrows).

Fig. 3 An 83-year-old male subject with perforated duodenal ulcer. The CXR shows hyperlucent liver sign (broken arrows) and dolphin sign (straight arrows).

Fig. 5 A 76-year-old male subject with pylorus perforation. The KUB shows doge cap sign (arrows).

parietal peritoneum appearing as a triangular radiolucency is called the triangle sign (Fig. 2) [5,7].

• • 1. RUQ signs
       1. Hyperlucent liver sign. On the supine radio- graphs, the blacker density of the large intraperitoneal free gas anterior to the ventral hepatic surface replacing the brightness of the hepatic shadow is the sign called hyperlu- cent liver sign (Fig. 3) [5].
       2. Anterior superior oval sign. This sign refers to a single or multiple oval, round, or pear-shaped gas bubbles projected over the liver shadow (Fig. 4) [5,12].
       3. Fissure for ligament teres sign. Visualization of the extrahepatic part of ligament teres on supine radiographs is called fissure for ligament teres sign (Fig. 2) [5,12-14].
       4. Doge cap sign. This sign refers to triangle-shaped free air accumulated in Morison pouch on supine abdominal films (Fig. 5) [5,15].

Fig. 4 A 68-year-old male subject with proximal Jejunum perforation due to ischemic bowel disease. The CXR shows anterior superior oval sign (broken arrows) and left anterior superior oval sign (straight arrows).

• • • 1. Hepatic edge sign. An oblong saucer or cigar- shaped collection of free air may be seen in the Subhepatic space with its long axis directed superomedially following the liver contour (Fig. 6) [5,8,16].
      2. Dolphin sign. The undersurface of the long costal muscle slips of the diaphragm that indented the adjacent air- filled space in the RUQ on supine films is a sign of pneumoperitoneum (Fig. 3) [17].

Fig. 6 A 77-year-old male subject with Billroth II anastomosis perforation. The KUB shows hepatic edge sign (broken arrows) and football sign (straight arrows).

Fig. 7 A 73-year-old male subject with perforated gastric ulcer. The KUB shows inverted V sign (broken arrows) and focal radiolucency (straight arrows).

• • 1. Peritoneal ligament-related signs
       1. Falciform ligament sign. The intraperitoneal free air may outline the falciform ligament, which is seen as a linear density situated longitudinally within the right upper abdomen (Fig. 1) [5,10,18,19].
       2. Inverted V sign. An inverted “V” may be seen over the pelvis on the supine films caused by the 2 lateral umbilical ligaments outlined by free air (Fig. 7) [5,20,21].

Fig. 8 A 72-year-old male subject with Perforated peptic ulcer. The KUB shows urachus sign (arrows).

Fig. 9 A 72-year-old female subject with perforated gastric ulcer. The CXR shows bilateral subphrenic radiolucency (arrows).

• • • 1. Urachus sign. When pneumoperitoneum occurs, the urachus may be seen as a thin midline linear structure in the lower abdomen from the umbilicus to the dome of the urinary bladder (Fig. 8) [5,22].
    1. Other signs
       1. Football sign. It refers to a large oval radi- olucency in the shape of an American football producing a sharp interface with the parietal peritoneum on a supine abdominal radiograph (Fig. 6) [23,24].
       2. Cupola sign. The cupola sign, also called sad- dlebag or moustache sign, is seen as an arcuate lucency overlying the lower thoracic spine and projecting caudad to the heart on supine radiograph (Fig. 1) [25].
       3. Left-sided anterior superior oval sign. We defined a single or multiple oval, round, or pear-shaped free air projected over left upper quadrant abdomen as “left- sided anterior superior oval sign” (Fig. 4).
       4. Subphrenic radiolucency. Radiolucency appeared beneath the diaphragm, either right or left side on the supine CXR, was defined as “subphrenic radiolucency” (Fig. 9).
       5. Focal radiolucency. The abnormal gas pattern presenting on the supine films without fitting any sign mentioned above was called “focal radiolucency” (Fig. 7).

The erect CXR (either standing or sitting position) and left decubitus abdominal radiographs were also evaluated for presence of free air in subphrenic area or nondependent part of right abdomen, respectively. The sensitivities of every sign of free air on either supine CXR or supine KUB were calculated. The overall detection rates of free air were also determined and compared among supine CXR, supine KUB, erect CXR, and left decubitus abdominal radio- graphs. The association between the radiographic signs of free air and the perforation sites of hollow organs and the difference between supine CXR and supine KUB were verified by ?2 test, respectively.

Results

Among the 250 patients of our study, 66 (26.4%) were females, and 184 (73.6%) were males. The mean age was

Table 1 Summary of the etiologies of the different perforation

sites

Perforation site

Etiology (No.)

Percentage

UGI (n = 228) Peptic ulcer disease (220) 96.5

Corrosive injury (2) 0.9

Billroth II anastomosis leakage (2) 0.9

Iatrogenic injury (2) 0.9

Foreign body ingestion (1) 0.5

Tumor rupture (1) 0.5

Small bowel intestinal obstruction (8) 47.1

(n = 17) Foreign body ingestion (3) 17.6

ischemic bowel disease (3) 17.6

Necrotizing inflammation (1) 5.9

blunt abdominal trauma (1) 5.9

Iatrogenic injury (1) 5.9

Colon (n = 5) Colon obstruction (1) 20

Chronic inflammation (1) 20

Tumor rupture (1) 20

Blunt abdominal trauma (1) 20

Iatrogenic injury (1) 20

Supine KUB signs	No.	Sensitivity (%)
No any sign	36	19.6
Anterior superior oval	81	44.0
Hyperlucent liver	56	30.4
Focal radiolucency	43	23.4
Rigler	36	19.6
Falciform ligament	20	10.9
Cupola	16	8.7
Football	7	3.8
Hepatic edge	7	3.8
Left-sided superior anterior oval	6	3.3
Triangle	5	2.7
Inverted V	5	2.7
Fissure for ligament teres	3	1.6
Dolphin	2	1.1
Doge cap	1	0.5
Urachus	0	0

67.1 years (range, 1-98 years). There were 228 patients

(91.2%) with UGI perforation, 17 patients (6.8%) with small bowel perforation, and 5 patients (2.0%) with colon perforation. The etiologies of the hollow organ perfora- tion were summarized in Table 1. Peptic ulcer disease was the most common cause of hollow organ perforation. There was no statistically significant association between

Table 2 Statistical association between free air signs on supine KUB and perforation sites

these radiographic free air signs and the perforation sites

(Table 2).

Table 3 Summary of the sensitivity of each free air signs on the supine KUB (n = 184)

The sensitivities of each free air signs on the 184 supine KUB and the 150 supine CXR were summarized in Tables

3 and 4. The anterior superior oval sign was the most common radiographic signs of free air on either supine KUB (44%) or supine CXR (34%). The second common sign on supine KUB (30.4%) and supine CXR (27.3%) was hyperlucent liver sign. Focal radiolucency was the third common sign on supine KUB (23.4%), but subphrenic radiolucency was the third common sign on supine CXR (21.3%).

Table 4 Summary of the sensitivity of each free air signs on the supine CXR (n = 150)

Supine KUB signs	Perforation	sites	P
(n = 184)	UGI	Small Colon
		bowel
Anterior superior	75 (92.6%)	6 (7.4%) 0	.400
oval (n = 81)
Hyperlucent liver	55 (98.2%)	1 (1.8%) 0	.082
(n = 56)
Focal radiolucency	41 (95.3%)	1 (2.3%) 1 (2.3%)	.350
(n = 43)
Rigler (n = 36)	34 (94.4%)	1 (2.8%) 1 (2.8%)	.350
Falciform ligament (n = 20)	19 (95.0%)	0 1 (5.0%)	.156
Cupola (n = 16)	14 (87.5%)	1 (6.3%) 1 (6.3%)	.311
Football (n = 7)	7 (100%)	0 0	1.000
Hepatic edge (n = 7)	7 (100%)	0 0	1.000
Left-sided superior	5 (83.3%)	1 (16.7%) 0	.425
anterior oval (n = 6)
Triangle (n = 5)	5 (100%)	0 0	1.000
Inverted V (n = 5)	5 (100%)	0 0	1.000
Fissure for ligament	3 (100%)	0 0	1.000
teres (n = 3)
Dolphin (n = 2)	2 (100%)	0 0	1.000

Supine CXR signs	No.	Sensitivity (%)
No any sign	32	21.3
Anterior superior oval	51	34.0
Hyperlucent liver	41	27.3
Subphrenic radiolucency	32	21.3
Cupola	31	20.7
Rigler	17	11.3
Falciform ligament	17	11.3
Dolphin	6	4.0
Hepatic edge	5	3.3
Focal radiolucency	5	3.3
Fissure for ligament teres	3	2.0
Left-sided superior anterior oval	2	1.3
Inverted V	0	0
Urachus	0	0
Doge cap	0	0
Football	0	0
Triangle	0	0

Categorized free air signs	Supine KUB (n = 184)	Supine CXR (n = 150)
RUQ	138 (75.0%)	95 (63.3%)
Others	68 (37.0%)	57 (38.0%)
Related to bowel wall	38 (27.7%)	17 (11.3%)
Related to peritoneal ligament	23 (12.5%)	17 (11.3%)

The overall sensitivities of each categorized group of free air signs on the 184 supine KUB and the 150 supine CXR were summarized in Table 5. Most free air signs on supine radiographs were located over the RUQ of the abdomen (75% on supine KUB and 63.3% on supine CXR). The positive rates of free air detection of the radiographic images with different positions such as supine KUB, supine CXR, erect CXR, and left decubitus abdominal radiograph were compared in Table 6.

Table 5 Summary of the overall sensitivity of categorized free air signs on the supine radiographs

There were 104 cases receiving both supine KUB and supine CXR simultaneously. The positive rate for free air was 87.5% on supine KUB and 80.8% on supine CXR. There was no statistically significant difference between each other. The same result was also noted between the positive rates for free air detection on erect CXR and left decubitus abdominal radiograph (84% vs 96%) in the 25 cases who undertook both radiographic images simultaneously.

Discussion

Gastrointestinal tract perforation can be caused by a variety of reasons, including peptic ulcer disease, trau- matic event, foreign body, appendicitis, Diverticular disease, inflammatory bowel disease, abscess, neoplasm, and iatrogenic factors [2,24]. In the literature, gastro- intestinal tract perforation has been reported to cause pneumoperitoneum from 40% to 80% of patients, and peptic ulcer diseases are the reasons of pneumoperitoneum in about 90% [2,6]. In our study, we found that the incidence of pneumoperitoneum with UGI perforation is 91.2%, and peptic ulcer diseases are the most common cause in 96.5% of these patients. Our findings are consistent with previous studies.

In our research, small bowel perforation resulted in 6.8% of pneumoperitoneum, and intestinal obstruction-induced perforation accounts for the most common etiology in approximately 47.1%. In previous report, the incidence of the small bowel perforation was 20% [6]. The perforation of colon only occurred in 2.0% in our study, whereas the incidence was reported as 34% in the literature [6]. We think that such a difference might be due to the small case number of our study. Perforated appendicitis and diverti- culitis of the colon are known 2 common reasons for the colon and rectum perforation [2,6]. The progressive

Inflammatory process provides very opportunity of sponta- neous sealing off of the perforation by surrounding Soft tissues. Therefore, it is uncommon for them to result in pneumoperitoneum [2].

In our study, anterior superior oval sign was the most common sign of pneumoperitoneum on the supine radio- graphic images, which was present in 44.0% on supine KUB and 34.0% on supine CXR. The second most frequent sign was hyperlucent liver sign, which appeared in 30.4% on supine KUB and 27.3% on supine CXR. We categorized the free air signs to 4 groups: bowel-related signs, RUQ signs, peritoneal ligament-related signs, and other signs. The overall sensitivity of the RUQ free air signs, including hyperlucent liver sign, anterior superior oval sign, fissure for ligament teres sign, doge cap sign, hepatic edge sign, and dolphin sign, was as high as 75.0% on supine KUB and 63.3% on supine CXR. It is obviously higher than that of the previous report (41%) [8]. We presume that such difference may be due to relative unfamiliarity of anterior superior oval sign that was relatively lately documented and not very noticeable in the radiologic literatures. The RUQ free air signs are the most frequent signs of pneumoperitoneum on supine radiographs. This location is the important place to look for small accumulations of intraperitoneal free gas [5,8]. We should pay attention to this area while reading the supine radiographic images. Our study showed no significant correlation between free air signs and perfora- tion sites (Table 2). Therefore, free air signs on the supine KUB cannot be used to predict the peroration sites of the gastrointestinal tract.

The overall positive rates of the free air signs in our study were 80.4% on supine KUB and 78.7% on supine CXR. It is higher than the reported data about supine radiographic images for detecting intraperitoneal free air, ranging from 56% to 59% in the literatures [5,8,16]. The higher sensitivity of free air detection was mainly attributed to the recognition of anterior superior oval sign. Familiarity with the various signs of intraperitoneal free air on supine radiographic images could lead to more accurate diagnosis of hollow organ perforation.

In the literatures, taking radiographic images with different positions can result in different sensitivities for the pneumoperitoneum [2,5-7]. Either erect CXR or left decubitus abdominal radiographs have been thought as the

Table 6 Summary of the positive rates of each radiographic images for free air detection

Radiographic films	No. of positive/ no. of films	Sensitivity (%)
Supine KUB	148/184	80.4
Supine CXR	118/150	78.7
Erect CXR	80/94	85.1
Decubitus radiograph	97/99	98.0

standard plain film projections to detect intraperitoneal free gas. The 2 kinds of radiographic images can demonstrate 20 to 30 mL of intraperitoneal free gas on average. They can even detect as little as 1 to 2 mL of intraperitoneal free gas in the right subphrenic area, where the upper most region of the abdominal cavity under a negative pressure after maintaining the upright position for 10 minutes or the Left lateral decubitus position for 10 to 20 minutes is normally located.

In our study, the sensitivity of erect CXR to detect free air is as high as 85.1%, as compared with the reported 46% to 80% in the literature [26,27]. The sensitivity appeared even higher (98.0%) on left decubitus abdominal radiographs in our study, which is similar to the report by Roh et al [6] (96.0%). These 2 kinds of radiographic images are more sensitive than supine KUB and supine CXR (Table 6).

There is no significant difference between the sensitivities of supine KUB and supine CXR for free air detection. However, the quality of supine CXR is always whiter than that of supine KUB, so that it might increase the difficulty for physicians to survey the pneumoperitoneum rapidly at ED, especially when there is a small amount of intraperitoneal free gas. In our experience, supine KUB is superior to supine CXR because the hollow organ perforation is highly suspected. Comparing erect CXR with left decubitus abdominal radiograph, it also revealed no significant difference between their sensitivities. The result means that the 2 kinds of radiographic films are equally effective in diagnosing hollow organ perforation.

Limitation

The proportions of small bowel and colon perforation are less in our study than those in previous literatures. It might be due to the small case number of our study. Besides, this is a retrospective study with small population. There might be inherent bias in interpreting these cases for free air and statistical analysis. The consensual determination of the presence or absence of free air sign may be influenced by knowledge of the purpose of this study, especially for minimal intraperitoneal free air. The imaging quality of the radiographs also influences the judgment of the viewers, such as overexposure or underexposure, inappropriate positioning, and inadequate field of view.

Conclusions

radiologic studies play an important role in the diagnosis of pneumoperitoneum. The demonstration of intraperitoneal free air on the radiographic images is often the initial clue to diagnose hollow organ perforation in an

ED. Familiarity with the free air signs is important to emergency physicians and radiologists. Most free air signs on supine radiographs are located over the right upper abdomen, with anterior superior oval sign as the most common sign. Supine CXR is as sensitive as supine KUB for free air detection. Because either supine KUB or supine CXR can detect most intraperitoneal free air, there is no need to take both images for all patients with suspected pneumoperitoneum. Only if supine radiograph is incon- clusive while hollow organ perforation is still suspected clinically, left decubitus abdominal radiograph or erect CXR should be taken.

References

1. Hainaux B, Agneessens E, Bertinotti R, et al. Accuracy of MDCT in predicting site of gastrointestinal tract perforation. Am J Roentgenol 2006;187:1179-83.
2. Ghahremani GG. Radiologic evaluation of suspected gastrointestinal perforation. Radiol Clin North Am 1993;31:1219-34.
3. Stapakis JC, Thickman D. Diagnosis of pneumoperitoneum: abdom- inal CT vs. upright chest film. J Comput Assist Tomogr 1992;16: 713-6.
4. Yeung KW, Chang MS, Hsiao CP, et al. CT evaluation of gastrointestinal tract perforation. Clin Imag 2004;28:329-33.
5. Cho KC, Baker SR. Extraluminal air diagnosis and significance. Radiol Clin North Am 1994;32:829-44.
6. Roh JJ, Thompson JS, Harned RK, et al. Value of pneumoper- itoneum in the diagnosis of Visceral perforation. Am J Surg 1983; 146:830-3.
7. Miller RE, Nelson SW. The roentgenological demonstration of tiny amounts of free intraperitoneal gas: experimental and clinical studies. Am J Roentgenol 1971;112:574-85.
8. Levine MS, Scheiner JD, Rubesin SE, et al. Diagnosis of pneumoperitoneum on supine abdominal radiographs. Am J Roent- genol 1991;156:731-5.
9. Ly JQ. The Rigler sign. Radiology 2003;228:706-7.
10. Miller RE. The roentgenological evaluation of intraperitoneal gas (pneumoperitoneum). CRC Crit Rev Clin Radiol Nucl Med 1973;4:61-85.
11. Rigler LG. Spontaneous pneumoperitoneum: a roentgenologic sign found in the supine position. Radiology 1941;37:604-7.
12. Cho KC, Baker SR, Thornhill BA, et al. Supine film diagnosis of pneumoperitoneum: new observations in the right upper quadrant. Radiology 1988;169(p):405.
13. Cho KC, Baker SR. Visualization of the extrahepatic segment of the ligamentum teres: a sign of free air on plain radiographs. Radiology 1997;202:651-4.
14. Cho KC, Baker SR. Air in the fissure for the ligamentum teres: new sign of intraperitoneal air on plain radiographs. Radiology 1991;178:489-92.
15. Brill PW, Olson SR, Winchester P. Neonatal necrotizing enterocolitis: air in Morison pouch. Radiology 1990;174:469-71.
16. Menuck L, Siemers PT. Pneumoperitoneum: importance of right upper quadrant features. Am J Roentgenol 1976;127:753.
17. Cho KC, Baker SR. Depiction of diaphragmatic muscle slips on supine radiographs: a sign of pneumoperitoneum. Radiology 1997;203:431-3.
18. Lin JN. Images in clinical medicine. Falciform-ligament sign of pneumoperitoneum. N Engl J Med 2004;351:e16.
19. Han SY. Variations in falciform ligament with pneumoperitoneum. Can Assoc Radiol J 1980;31:171-3.
20. Bray JF. The “inverted V” sign of pneumoperitoneum. Radiology 1984;151:45-6.
21. Weiner CI, Diaconis JN, Dennis JM. The inverted V: a new sign of pneumoperitoneum. Radiology 1973;107:47-8.
22. Jelasco DV, Schultz EH. The urachus: an aid to the diagnosis of pneumoperitoneum. Radiology 1969;92:295-329.
23. Rampton JW. The football sign. Radiology 2004;231:81-2.
24. Nelson SW. Extraluminal gas collections due to diseases of the gastrointestinal tract. Am J Roentgenol 1972;115:225-48.
25. Marshall GB. The Cupola sign. Radiology 2006;241:623-4.
26. Chen CH, Yang CC, Yen YH. Role of upright chest radiography and ultrasonography in demonstrating free air of perforated peptic ulcers. Hepatogastroenterol 2001;48:1082-4.
27. Woodring JH, Heiser MJ. Detection of pneumoperitoneum on chest radiographs: comparison of upright lateral and posteroranterior projections. Am J Roentgenol 1995;165:45-7.