British Journal of Radiology (2003) 76, 437-443
© 2003 British Institute of Radiology
doi: 10.1259/bjr/16438547
The aurora sign: an ultrasonographic sign suggesting parenchymal lung disease
S Kohzaki, MD1,
K Tsurusaki, MD2,
M Uetani, MD2,
K Nakanishi, MD2 and
K Hayashi, MD2
1 Department of Radiology, Kitakyushu City Yahata Hospital, 4-18-1 Nishihon-machi, Yahatahigashi, Kitakyushu 850-8534 and 2 Department of Radiology, Nagasaki University School of Medicine, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan
 |
Abstract
|
|---|
The objective of this study was to clarify the cause and clinical significance of a large number of ring-down artefact (RA) observed on the dorsal side of the right hepatic lobe on abdominal ultrasound (US). 2000 abdominal US examinations were evaluated to investigate the frequency and number of RA behind the right lobe of the liver. In this study, RA observed by subcostal or intercostal US were described as the "aurora sign" when they were numerous. US findings were correlated with high resolution CT or three-dimensional CT of the right lung base. Experimental study was also performed to investigate the mechanism of the aurora sign. The results were as follows. (1) Aurora sign was noted in 43 patients. In 37 of these 43 patients (86.0%), chest CT or plain radiography revealed diffuse interstitial changes in the right lower lung field. Three-dimensional CT of the lung and the experimental model revealed that the aurora sign derived from the irregularity of air spaces immediately below the pleura associated with interstitial pulmonary disorders. (2) One or more bands of RA were observed in 907 of the 2000 patients (45.4%). Of 177 patients with fatty liver, RA was observed in 14 (7.9%), while of the 1823 patients with no fatty liver, RA was observed in 893 (49.0%). The frequency of RA was significantly different (p<0.001) between the patients with and without fatty liver. In conclusion, parenchymal lung disease should be suspected when the aurora sign is noted on the dorsal side of the right hepatic lobe. However, RAs are rare in patients with fatty liver.
|
Introduction
|
|---|
It has long been known that multiple long vertical bands or lines of high echo are occasionally observed to arise from the dorsal aspect of right hemidiaphragm. They are variously explained as reverberation artefact of unknown cause [1], reverberation artefact due to intrapulmonary gas [2], or as comet tail artefact of the diaphragm [3], but the true mechanism of their formation has not been clarified.
Reverberation artefact is a phenomenon produced by repeated reflections of ultrasound between a certain distance. The artefact often occurs between the transducer and an almost total reflector such as air or bone. In the human body, intramural gallstone, intrahepatic calcification, Rokitansky–Aschoff sinus in adenomyomatosis of the gallbladder, a bullet or metallic clip may cause reverberation artefact. In these small reflective objects, multiple reflections are produced between their proximal and distal surfaces, and strong echoes simulating comet tail are observed: it has thus been termed comet tail artefact [4].
An artefact that closely simulates this comet tail artefact also arises from internal gas as seen in pneumobilia or pneumoretroperitoneum. This has also been attributed to reverberation. Avruch et al investigated this phenomenon conducting an experimental study. According to them, this is not due to reverberation but due to resonance of ultrasound that occurs in a cavity surrounded by small air bubbles (air tetrahedron), and has been termed a ring-down artefact (RA) [5].
We previously reported the occurrence of a large number of RAs seen in pneumatosis cystoides intestinalis and named this unique ultrasound image the aurora sign [6]. We have since noted the frequent appearance of ultrasound images closely resembling the aurora sign on the dorsal side of the right hepatic lobe in patients with no abdominal abnormalities [7]. Therefore, we prospectively examined the frequency of the occurrence of RAs on the dorsal side of the right hepatic lobe, their number, and their relationships with underlying diseases to clarify their cause and clinical significance. Experimental study was also performed to investigate the mechanism of the aurora sign.
|
Methods and materials
|
|---|
Patient study
The subjects were 2000 consecutive patients (1101 males and 899 females) who underwent abdominal ultrasound (US) between June 1994 and January 1996. They aged from 0 years to 89 years (mean 54.0 years). Patients with a history of right hepatic lobectomy or thoracotomy, those with right pleural effusion, those with right subdiaphragmatic ascites, and markedly obese patients were excluded.
In this study, high echo bands 3–10 mm wide and 30 mm or greater in length were defined as RA, and their frequency of occurrence and number were examined prospectively by subcostal and intercostal US (Figure 1
). Subcostal and intercostal imaging were performed, each in one deep inspiration, and the images were hardcopied whenever RAs were found. The number of RAs was calculated by summing up the number of them observed with both imaging techniques.
View larger version (101K):
[in this window]
[in a new window]
|
Figure 1. Hepatic ultrasounds obtained from a patient without pulmonary disease. (a) Subcostal and (b) intercostal ultrasound of the liver show some ring-down artefacts that originate from the right diaphragm with perpendicular orientation.
|
|
Since our daily experience suggested that RA was less frequently observed in patients with fatty liver, differences between patients with and without fatty liver were examined. The diagnosis of fatty liver was based on US findings of an increased echo level in the hepatic parenchyma, enhanced hepatorenal contrast, obscuration of intrahepatic vessels, and attenuation of deep echoes.
In patients who underwent thoracic CT 3 months before or after US (184 of the 2000 patients), findings of the two examinations were compared.
RAs observed by subcostal or intercostal US were defined as the aurora sign when they were numerous (more than 20). When the aurora sign was observed, chest CT was obtained as soon as possible if not already taken. From June 1994 to March 1995, oblique axial CT, which was performed by slanting the gantry 25° to the caudal direction, was also carried out to obtain images close to the view of subcostal US. After April 1995, when helical CT scan became available, three-dimensional (3D)-CT of the right lower lung field was obtained, substituting for the oblique axial CT. 3D-CT of the right lower lung field was also obtained from 10 healthy volunteers with no lung abnormality (see postscript 2).
US examination was performed using an Aloka SSD670 (Tokyo, Japan) ultrasound machine with 3.5 MHz convex type and 3.5 MHz sector type probes for adult patients and with 5.0 MHz sector type probe for paediatric patients. All examinations were performed by two experienced radiologists (SK, KN) and one resident (KT). Evaluation was made directly on screen during the examination and hard-copy pictures were taken.
CT was performed using a Toshiba TCT900S (Tokyo, Japan) (between June 1994 and March 1995) or a GE Hi-Speed Advantage SP (Milwaukee, WI) (between April 1995 and January 1996). The slice thickness and slice interval were 7 mm and 7 mm, respectively, for conventional thoracic CT; 1.5 mm and 1.5 mm, respectively, for high resolution (HR)-CT; and 3 mm and 3 mm, respectively, for oblique axial CT. 3D-CT images were obtained by scanning the right lung base by helical CT over 15 cm in the craniocaudal direction during a single breath hold, and reconstructing 3D images (5 mm section thickness, 1:1 pitch, image reconstruction: GE Advantage Windows-Sdc, V1.2.6, reconstructed at 1 mm interval by standard algorithm, low threshold -1000 HU, high threshold -700 HU).
The diagnosis of lung disease was made after reviewing the medical records in all 2000 patients, and after the discussion with attending physicians based on the patient's symptoms, histories, CT findings and in some patients, pathological findings.
Experimental study
We conducted an experimental study to investigate the mechanism of RA. A sheet of vinyl packing material with a honeycomb pattern of air-filled capsules (each measuring 8 mm in diameter) was attached to the bottom of a nylon bag filled with water and was scanned from within the bag with the 3.5 MHz convex probe.
|
Results
|
|---|
Frequency and number of RA
One or more RA were observed in 907 (45.4%) of the 2000 patients. The aurora sign (numerous RA) was noted in 43 patients (2.2%). One to 14 (mean 2.9) bands of RA were observed in 864 patients (43.2%) (Table 1).
Effects of fatty liver on frequency of RA
The diagnosis of fatty liver was made in 177 of the 2000 patients. Of these 177 patients, RA was observed in 14 (7.9%). Of the 1823 patients without fatty liver, RA was observed in 893 (49.0%). The frequency of RA significantly differed (p<0.001) between the patients with and without fatty liver (Table 2).
Analysis of 43 patients with aurora sign
The aurora sign was noted by subcostal or intercostal US in 43 patients (Table 3) (Figures 2 and 3). CT was performed in 39 of them, excluding 2 neonates and 2 adults. Diffuse interstitial changes were demonstrated in the right lower lung field in 35 of them. The two adults who did not undergo CT also showed marked interstitial changes in the middle and lower lung fields on chest radiograph. Of these 37 patients with interstitial changes, 24 had idiopathic pulmonary fibrosis (IPF), 6 had emphysema, 5 had collagen lung (diffuse parenchymal lung disease associated with collagen vascular disease), 1 had lymphoid interstitial pneumonia (LIP), and 1 had cystic bronchiectasis. Of the four patients without interstitial changes, one had right apical pneumothorax, one had atalectasis of the left lower lobe, and two were intensely positive for C-reactive protein.
View larger version (70K):
[in this window]
[in a new window]
|
Figure 2. A 71-year-old man with idiopathic pulmonary fibrosis. (a) Subcostal ultrasound of the liver shows numerous ring-down artefacts. (b) Chest CT scan shows honeycomb changes of the right lung base.
|
|
View larger version (70K):
[in this window]
[in a new window]
|
Figure 3. A 65-year-old woman with collagen lung. (a) Intercostal ultrasound of the liver shows numerous ring-down artefacts. (b) Chest CT scan did not show any honeycomb changes but did show increased interstitial changes of the parenchyma of the right lung base.
|
|
The aurora sign had disappeared in the 2 neonates by repeat US performed after about 1 week. It had also disappeared in the four adults without interstitial changes on US performed after alleviation of the symptoms.
Negative aurora sign in patients with parenchymal lung disease
Diffuse interstitial changes were noted in the right lower lung field in 42 of 184 patients who underwent thoracic CT within 3 months before or after US (Table 4). The aurora sign was positive in 35 of the 42 patients and negative in the remaining 7. Observed abnormalities in those 7 patients included fatty liver in 5, dense pleural calcification at the lung base in 1, and Chilaiditi syndrome in 1.
View this table:
[in this window]
[in a new window]
|
Table 4. Absence of the aurora sign in the patient with diffuse interstitial change in the right lower lung (n=7)
|
|
Findings on oblique axial CT
Interstitial changes of the right lower lung field were noted in all 23 patients who underwent oblique axial CT. Honeycomb changes extending along the right hemidiaphragm were clearly visualized in 13 of them (Figure 4). In the remaining 10 patients, increased lung density or thickening of bronchovascular bundle or bullous emphysema was observed.
View larger version (72K):
[in this window]
[in a new window]
|
Figure 4. (a) Scout image of the oblique axial scan. (b) Oblique axial CT of the lung base performed by slanting the gantry 25 degrees in the caudal direction clearly showed honeycomb changes extending along the right hemidiaphragm.
|
|
Comparison with 3D-CT
3D-CT in 10 aurora sign-positive patients with parenchymal lung diseases exhibited marked irregularity or cobble stone appearance of the lung surface, while that in 10 healthy subjects exhibited smooth surfaces except at the interlobar fissure and at the region surrounding the inferior vena cava (Figure 5).
View larger version (79K):
[in this window]
[in a new window]
|
Figure 5. (a) Three-dimensional CT of the right lung base obtained from a healthy subject showed smooth surfaces except for the interlobar fissure and the region surrounding the inferior vena cava. (b) Three-dimensional CT of the right lung base obtained from an aurora sign-positive patient showed marked irregularity or cobble stone appearance of the lung surface.
|
|
Findings of experimental study
When a sheet of vinyl packing material with a honeycomb pattern of air-filled capsules was attached to the bottom of a nylon bag filled with water, a large number of RAs was observed (Figure 6).
View larger version (54K):
[in this window]
[in a new window]
|
Figure 6. (a) The schema of an experimental model and an "aurora sign". (b) Experimental ultrasound. The scanner is submerged in the water and a sheet of vinyl packing material is attached outside the water bag, displaying an "aurora-like" image.
|
|
When only one air-filled capsule was attached to the bottom and scanned, no RA was observed. When two or three of the air-filled capsules were scanned, one or two RAs were observed, respectively. In addition, the ultrasound images, obtained after inserting a small marker at the top of the capsule, confirmed that RA derived from between the two capsules.
|
Discussion
|
|---|
Systemic investigation of RA behind the right hepatic lobe has only been carried out by Lim et al [8]. They reported that respiratory disorders should be suspected when numerous RA bands are observed on the dorsal side of the right hepatic lobe by abdominal US, even if there are no respiratory signs.
Suramo et al classified acoustic shadows caused by gallstones, intrahepatic stones, and foreign bodies and reverberation artefacts into several patterns and compared them with underlying diseases [1]. They stated that RA was occasionally observed on the dorsal side of the right hemidiaphragm but the origin was unknown. Lewandowski et al studied mirror image artefacts of the diaphragm and reported that RA due to pulmonary gas was observed in 38 (47.5%) of 80 patients [2]. In the present study, RA was observed in 907 (45.4%) of the 2000 subjects, and the frequency was similar to that reported by Lewandowski et al.
The frequency of RA was clearly lower in patients with fatty liver (7.9%) than in those without fatty liver (49.0%), probably because of attenuation of US due to fatty liver. The frequency and number of RA may differ according to the US frequency of the probe, although this was not evaluated in this study.
Since the aurora sign that appears on the dorsal side of the right hemidiaphragm closely resembles that observed in patients with pneumatosis cystoides intestinalis, these conditions are considered to be due to the same mechanism. We previously proposed a mechanism of aurora sign in pneumatosis cystoides intestinalis [6]. However, there was a fundamental misunderstanding; we thought bright echoes originated from the dorsal aspects of the air-filled capsules. We thus stated "the bright echoes from the air-filled capsules may be reverberation artefacts created between the anterior and posterior walls of the capsules". From our repeat experiment, we confirmed that the bright echoes had their origins between the air-filled capsules. Figure 6
shows the results of our experimental study. When vinyl-packing material with many air capsules was scanned by the water immersion method, acoustic shadows were produced at the air capsules, and high echo bands appeared in areas between the air capsules. Furthermore, there was no echo when only one air-filled capsule was attached to the bottom and scanned, while one or two high echo bands were observed when two or three of air-filled capsules were scanned, respectively. These high echo bands are considered to be the same as "ring-down artefact" reported by Avruch et al [5].
The RA appears as a line in the direction of the sound beam and deep to a strong reflector. The cause can be a piece of metal in the body, such as a surgical clip or lead shot. More commonly, the RA is seen deep to a collection of gas. In fact, the artefact is not an echogenic line but rather a collection of closely spaced perpendicular echoes along one or two vectors in the US image. Originally, they were thought to be the sound entering the metal or gas bubble and reverberating back and forth within the structure, each time sending some of the sound back to the transducer. This explanation would account for a series of closely spaced echoes equal to the depth of the piece of metal or gas bubble. A more plausible explanation, however, is equivalent to the ringing of a bell. The sound pulse insonates the metal, causing it to ring. With gas bubbles, there is a slightly different explanation. Not all collections of gas produce RA, but the RA can be characteristic of certain configurations of gas bubbles. The "bugle" of fluid trapped between four small bubbles (the bubble tetrahedron), is the source of the artefact. The sound pulse insonates the bugle, causing it to vibrate and send a prolonged sound wave back to the transducer [9].
From these findings, a large number of high echo bands observed on the dorsal side of the right hepatic lobe are considered to be derived from the irregularity of air spaces immediately below the pleura associated with interstitial pulmonary disorders. This view is supported by the findings that interstitial changes were observed in the lung parenchyma adjacent to the diaphragm in all patients showing the aurora sign on oblique axial CT and that the cobble stone appearance was observed in these patients on 3D-CT.
On the other hand, in four cases with aurora sign on US, chest CT did not show interstitial changes. In two of them, C-reactive protein was positive and there were respiratory symptoms. On repeat US after C-reactive protein became negative and respiratory symptoms subsided, the aurora sign had disappeared. A subtle finding of pulmonary inflammatory process may have caused the aurora sign in these cases. In a case with right pneumothorax and a case with left lower lobe collapse, the aurora sign was no longer present in repeat US after the disappearance of the lesions. These cases suggest that the pulmonary lesions other than interstitial disease may also cause the aurora sign.
Two neonates were included in this study. In both cases, there was no definite pulmonary abnormality but US demonstrated aurora sign 1 day after birth that disappeared 7 days later. The normally aerated lung is highly reflective and presents barriers to sound transmission. However, it has been shown that when alveolar air is replaced by fluid the lung becomes hypoechoic with good sound transmission [10–12]. Avni and his colleague reported that "retrohepatic diffuse hyperechogenicity" replacing the normal diaphragm complex in newborns indicates hyaline membrane disease and this resolved (within a mean time of 6.5±5 days) completely in patients with an uncomplicated clinical evolution [13, 14]. In contrast, in patients with chronic lung disease, the hyperechogenicity was only partially resolved, resulting in less diffuse and less extensive hyperechogenicity. The origin of "retrohepatic diffuse hyperechogenicity" has not been completely clarified. They speculated that the summation of many airways surrounded by totally collapsed alveoli might have induced retrohepatic hyperechogenicity. US images of some cases presented as retrohepatic diffuse hyperechogenicity in these articles mimic the aurora sign but the echogenic areas are more diffuse. This is probably because the air-trapped capsules are small and produce RA that are thinner and denser. We consider that retrohepatic diffuse hyperechogenicity described by Avni et al and the aurora sign may be the same phenomenon.
Interstitial changes were noted, but no aurora sign was observed, in 7 patients, namely 5 with fatty liver, 1 with pleural calcification, and 1 with Chilaiditi syndrome, due to attenuation of US by the fatty liver, calcification, and the intestine, respectively. The aurora sign was not observed in 8 patients who showed a localized increase in lung density on the dorsal side of the right lower lung field or in 2 patients with emphysema, probably because the lesions were not close to the diaphragm.
In conclusion, parenchymal lung disease should be suspected when the aurora sign is noted on the dorsal side of the right hepatic lobe. However, RAs are rare in patients with fatty liver.
Postscript 1: The first author (SK) was a naval surgeon on the Japan Maritime Self-Defense Force Icebreaker "Shirase" between 1988 and 1989. The aurora at the Antarctic Circle gave him the idea for this work.
Postscript 2: On editorial enquiry, the authors stated the following: "we obtained 3D-CT of the right lower lung field in 10 healthy volunteers in this study, we had not obtained the approval of the relevant ethics committee; this was not required under Japanese regulations at that time.
|
Acknowledgments
|
|---|
We would like to express our thanks to Yoshihiro Hayashida for the illustrations, and to Terumi Kanaya and Ikuko Uematsu for their valuable assistance. The technical support of Fumio Tagawa is also gratefully acknowledged.
Received for publication July 25, 2002.
Revision received December 23, 2002.
Accepted for publication April 9, 2003.
|
References
|
|---|
- Suramo I, Paivansalo M, Vuoria P. Shadowing and reverberation artifacts in abdominal ultrasonography. Eur J Radiol 1985;5:147–51.[Medline]
- Lewandowski BJ, Winsberg F. Echogenic appearance of the right hemidiaphragm. J Ultrasound Med 1983;2:243–9.[Abstract]
- Kremkau FW, Taylor KJW. Artifacts in ultrasound imaging. J Ultrasound Med 1986;5:227–37.[Abstract]
- Weill FS. Ultrasound diagnosis of digestive diseases, 3rd rev. edn. Berlin Heidelberg: Springer-Verlag, 1990:26–36.
- Avruch L, Cooperberg PL. The ring-down artifact. J Ultrasound Med 1985;4:21–8.[Abstract]
- Kohzaki S, Hayashi K, Fukuda T, et al. The "aurora sign"—a new sonographic sign of pneumatosis cystoides intestinalis. Br J Radiol 1994;67:1275–7.[Abstract]
- Kohzaki S, Tsurusaki K, Uetani M, Nakanishi K, Hayashi K. The aurora sign: a sign of interstitial lung disease on abdominal ultrasonography. Ultrasound Med Biol 1997;23:S30.
- Lim JH, Lee KS, Kim TS, Chung MP. Ring-Down artifacts posterior to the right hemidiaphragm on abdominal sonography: sign of pulmonary parenchymal abnormalities. J Ultrasound Med 1999;18:403–10.[Abstract]
- Muglia V, Cooperberg PL. Ring-down artifacts. In: Mcgahan JP, Goldberg BB. Diagnostic ultrasound — a logical approach (1st edn) Philadelphia, PA: Lippincott-Raven 1998;24–6.
- Targhetta R, Bourgeois JM, Chavagneux R, et al. Ultrasound signs of pneumothorax: Preliminary work. J Clin Ultrasound 1993;21:245–50.[Medline]
- Weinberg B, Diakoumakis EE, Kass EG, Seife B, Benzvi Z. The air bronchogram. Sonographic demonstration. AJR Am J Roentgenol 1986;147:593–5.[Abstract/Free Full Text]
- Cosgrove DO, Garbitt P, Hill CR. Echoes across the diaphragm. Ultrasound Med Biol 1978;3:385–92.[CrossRef][Medline]
- Avni EF, Braude P, Pardou A, Matos C. Hyaline membrane disease in the newborn: diagnosis by ultrasound. Pediatr Radiol 1990;20:143–6.[CrossRef][Medline]
- Avni EF, Cassart M, Maertelaer V, Rypens F, Vermeylen D, Gevenois PA. Sonographic prediction of chronic lung disease in the premature undergoing mechanical ventilation. Pediatr Radiol 1996;26:463–9.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
J. Huang, J. K. Triedman, N. V. Vasilyev, Y. Suematsu, R. O. Cleveland, and P. E. Dupont
Imaging Artifacts of Medical Instruments in Ultrasound-Guided Interventions
J. Ultrasound Med.,
October 1, 2007;
26(10):
1303 - 1322.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C-L Tsai, H-P Wang, W-C Lien, C-C Chen, T-I Lai, and W-J Chen
Ring down artefacts on abdominal sonography to predict pulmonary abnormalities in the emergency department
Emerg. Med. J.,
October 1, 2005;
22(10):
747 - 748.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
