Ultrasound in Med. & Biol., Vol. 32, No. 8, pp. 1157–1163, 2006 Copyright © 2006 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/06/$–see front matter doi:10.1016/j.ultrasmedbio.2006.04.006

● Original Contribution

THE ULTRASONOGRAPHIC DEEP SULCUS SIGN IN TRAUMATIC PNEUMOTHORAX

† † † GINO SOLDATI,* AMERICO TESTA, GIULIA PIGNATARO, GRAZIA PORTALE, † ‡ † DANIELE G. BIASUCCI, ANTONIO LEONE, and NICOLÒ GENTILONI SILVERI *Operative Unit of Emergency Medicine, Castelnuovo Garfagnana Hospital, Lucca, Italy; and †Department of Emergency Medicine and ‡Department of Radiology, A. Gemelli University Hospital, Rome, Italy

(Received 11 November 2005; revised 17 March 2006; in final form 4 April 2006) Abstract—A series of 186 patients with blunt chest trauma was studied with transthoracic ultrasonography to diagnose pneumothorax and to evaluate its size and location. The results were compared with bedside chest radiography and spiral CT scan. The prevalence of pneumothorax on CT scan was 56/186 (30.1%). Pneumo- thorax was proven on radiography in 30/56 cases without false positive results: “radiographic deep sulcus sign” was evident in 3/29 cases, 26/29 cases being occult. The ultrasound study demonstrated the presence of pneumothorax in 55/56 patients: one occult pneumothorax was missed and no false positive results were observed. The CT scan differed of ؎2.3 cm (range 1–5 cm) from the US study in evaluating size and location of pneumothorax. In conclusion, ultrasound study may detect occult pneumothorax undiagnosed by standard plain radiography. It reflects accurately the extent of pneumothorax if compared with CT scan, outlining the “ultrasonographic deep sulcus sign” on anterior chest wall. (E-mail: [email protected]) © 2006 World Federation for Ultrasound in Medicine & Biology. Key Words: Chest ultrasonography, Transthoracic ultrasonography, Chest trauma, Pneumothorax, Deep sulcus sign.

INTRODUCTION In the emergency room (ER), a few PNXs (7 to 30%) may be missed on conventional radiography in the Traumatic pneumothorax (PNX) is the most common life supine patient, being detected only on CT imaging (oc- threatening outcome in blunt chest trauma, occurring in cult PNXs) (Tocino et al. 1984). Some clues may suggest almost 40% of patients with penetrating chest injuries the presence of a PNX on chest plain radiographs: in a and over 20% of subjects with blunt injuries (Di Bar- supine patient, air collecting anteriorly within the tolomeo et al. 2001; LoCicero and Mattox 1989). Chest tube placement should be considered because even a hemithorax’s lower part causes a hyperlucent focus near small PNX can rapidly enlarge and become clinically the diaphragm and along the juxtacardiac region: this is significant, especially during general anaesthesia, with the “radiographic deep sulcus sign” (Fig. 1)(Gordon positive pressure ventilation in the setting of PEEP 1980). For the same reason, an improved sharpness of (Wescott and Cole 1983). the cardiomediastinal border is sometimes visible, but it Diagnosis of a PNX on standard chest radiography can be frequently overlooked. is based on the identification of a hyperlucent air space Static and dynamic ultrasound (US) features of between the visceral pleural layer, visible as a thin cur- PNX have been identified in a number of studies (Lich- vilinear white line marking the boundary of the lung and tenstein and Menu 1995; Rowan et al. 2002; Soldati and the parietal pleura layer. The findings of mediastinal shift Rossi 2000) and recently reviewed by Lichtenstein and diaphragmatic depression are the most important (2005). The technique involves the identification of the radiographic signs in the detection of tension PNX (Go- pleural line on conventional 2-D ultrasonography: the bien et al. 1982). observation of its mobility with respiration (“sliding sign”) and/or the presence of vertical artifacts which arise from the pleural line (comet-tails) ruling out the Address correspondence to: Dr. Americo Testa, Via dei Laghi PNX occurrence (Fig. 2), in which a completely hori- 32, 00040 Rocca di Papa, Roma, Italy. E-mail: [email protected] zontal line pattern (“A-line sign”) is shown (Fig. 3). The

1157 1158 Ultrasound in Medicine and Biology Volume 32, Number 8, 2006

Fig. 1. Anteroposterior supine chest radiograph, revealing the Fig. 3. US bidimensional (2-D) image in case of PNX. Over the anteromedial and inferior collection of a little amount of air in abolition of the sliding sign (dynamic sign), valuable just in real the antideclivous pleural space on the right side (small PNX) by time, the static sign of the replacement of the lung US back- the hyperlucency of the cardiomediastinal right border (black ground with an acoustic shadow exclusively containing several arrows) and the diaphragmatic dome (white arrows) on the horizontal reverberations (empty arrows) can be seen (A-lines same side. A severe PNX with diaphragmatic depression and sign). Vertical artifacts (comet-tails) arising from pleural line lung collapse without mediastinal shift was present on the left (full arrow) are absent. side.

demonstration of the lung sliding during respiratory mo- tion may also be shown by M-mode imaging (Soldati 2001) and, finally, enhanced by colour-power-Doppler mode (“power slide”) (Cunnigham et al. 2002). As the air is moving free, the US findings of PNX should be sought in the highest anterior area of the hemithorax’s lower part, near the diaphragm and the cardiomediastinic borders in the supine position. On the other hand, an accurate quick assessment of the PNX extent is mandatory in emergency. Therefore, it is essen- tial to identify the pleural contact reappearance points, the site on the chest wall in which a normal lung US pattern is replaced by a PNX US pattern (Lichtenstein et al. 2000). The aim of this study was to validate the transtho- racic ultrasonography in defining the presence, location and size of traumatic PNX in comparison with antero- posterior supine chest radiography and CT scan (gold standard) in emergency room. Furthermore, we will ex- plain a new US sign (“ultrasonographic deep sulcus Fig. 2. Ultrasound bidimensional (2-D) image of normal chest sign”), which we believe to be specific for an early and on longitudinal parasternal scan. The superficial layers are visible at the top of the screen. The ribs are recognised by their quick detection of PNX in a supine patient. curved shape producing posterior acoustic shadow. Below the curved rib line, the pleural line (full arrow) appears as an hyperechoic horizontal line, representing the lung surface-chest MATERIALS AND METHODS wall interface. A bat flying toward us can be imagined (the bat Patients sign). Vertical artifacts (empty arrow) appearing as hypere- choic beams arising from the pleural line and spreading up to A series of 186 subjects with blunt chest trauma (69 the edge of the screen across the lung field (comet-tails) can be women and 117 men, mean age 52.4 Ϯ 22.9 y, range produced. 16–89 y), consecutively admitted in the ER of Lucca and Pneumothorax echo-pattern G G. SOLDATI et al. 1159

Valle del Serchio General Hospitals and A. Gemelli University Hospital of Rome from 1 September 1999 to 31 August 2004, were examined for PNX with chest ultrasonography in a prospective blinded study. The breakdown of injuries was as follows: 114 motor vehicle crashes, 32 cycle and motor cycle crashes, 18 falls and 22 other injuries. Of 204 patients at first included in the study, five cases were unable to give informed consent and 11 cases requiring immediate operative interventions or with chest wall injuries precluding US evaluation (pain, ribs fractures, flail chest) were excluded; subcuta- neous emphysema hampered to obtain adequate US im- aging in two cases.

US examination The US examination was carried out within 30 min of admission, after primary survey, by emergency phy- sicians (EPs) not directly involving in the patient’s man- agement without hampering or delaying it. Subsequently, all patients underwent a plain chest radiograph and a spiral CT scan, which were all blindly read by radiolo- gists. The PNXs missed on conventional supine chest radiographs were defined “occult”. PNXs were classified on CT images, according to Wolfman et al. (1993), as “anterior” (anterior pleural air not extending to the mid- dle-axillary line dividing the thorax into equal anterior and posterior halves) or “anterolateral” (pleural air ex- tending at least to the middle-axillary line). Patient’s demographic data, examination’s time and findings were recorded on a data sheet for later evaluation, after having Fig. 4. Anatomical projection of the “deep sulcus sign” on obtained the patient’s informed consent. anterior chest wall is reported. The correct position of probe A Toshiba SSA-250A, an Esaote MEGAS-GP or a during the examination is indicated. The first look is conducted Hitachi H21 ultrasound machine, each equipped with a 5 by moving the probe in a longitudinal scan starting from the parasternal line: this approach provides to detect or rule out MHz convex array transducer, was used for US exami- PNX. The second look is conducted by directly exploring the nation. The choice of the 5 MHz convex array instead of intercostal spaces accurately to recognise the lung point: its aim the higher frequency linear transducer provides a good is to evaluate the size and location of PNX and to delimit the compromise between resolution and penetration, as pre- deep sulcus area. viously experienced (Soldati 2001; Soldati and Rossi 2000). The 5 MHz convex transducer is able to afford a complete visualization of the various horizontal and ver- nation in the parasternal line on both sides, from the tical artifacts mostly constituting the normal lung US third intercostal space to the diaphragm. When it was pattern and a simultaneous good penetration to the possible, the examination was begun in the apparently deeper layers. Moreover, the curved surface of the con- unaffected lung to determine a baseline. To locate the vex probe facilitates the scanning through intercostal lung surface, the probe was applied longitudinally spaces and it allows rapidly to perform the focused over an intercostal space to identify two contiguous assessment with ultrasonography for trauma (FAST) ribs: between these ribs, the pleural line, a hypere- with the extension to chest for ruling out PNX (E-FAST) choic horizontal band lying in a deeper plane, dis- without changing the transducer. Trained EPs with over played the so called “bat sign” (Lichtenstein 2005) 10 y experience in emergency ultrasonography and at (Fig. 2). If PNX US pattern occurred (as described least 1 y in chest ultrasonography have contributed to below), a second look was then systematically as- this study (GS, AT). sessed with transverse scans along the intercostal All subjects were longitudinally US scanned in spaces from sternal to middle-axillary line to evaluate the supine position with a two-look approach (Fig. 4). its size and location (Fig. 4)(Lichtenstein and Menu The first look consisted of anterior chest wall exami- 1995; Rowan et al. 2002; Soldati and Rossi 2000). 1160 Ultrasound in Medicine and Biology Volume 32, Number 8, 2006

sliding sign and comet tails: the “A-line sign”) (Fig. 5). Therefore, our experience suggests that the lung point efficiently marks the periphery of the pleural air collec- tion in the point where the visceral pleural layer restores its contact with the parietal pleural layer. The sonographic findings of PNX in the “deep sul- cus sign” area are what we have defined as the “ultra- sonographic deep sulcus sign”. The “deep sulcus area” is the anterior thoracic area corresponding to the projection of the cardiomediastinal borders and anterior costo- phrenic sulcus, bounded superiorly by the second rib, inferiorly by the diaphragm and laterally by the mam- mary lines above and the anterior axillary lines below (Fig. 4). The definitions of the various transthoracic ultrasonographic findings are summarized in Table 1 (Lichtenstein 2005; Lichtenstein and Menu 1995; Rowan et al. 2002; Soldati and Rossi 2000). Fig. 5. The lung point (full arrow): it represents the regular reappearance of lung sliding sign on the right side (without comet-tail artifacts in this picture), replacing a PNX 2-D US Statistics pattern on the left side (absent lung sliding and comet- Continuously distributed variables were expressed tail artifacts replaced by horizontal reverberation lines) (empty arrows). by mean values with SDs and categorised variables in contingency tables. Estimates of specificity, sensitivity and overall accuracy were calculated for radiography versus CT (assumed as reference standard for PNX de- US appearances tection), US versus CT and US versus radiography on The US appearance that was sought to assess PNX 372 lung fields ( ϭ 186 ϫ two patients). All data were diagnosis was the absence of the sliding sign, as well as analyzed with SPSS for Windows, version 10.0 (Chi- the detection of the comet-tail artifacts, which was con- cago, IL, USA) software. sidered enough to rule out PNX. The sliding sign de- scribes the movement of the pleural line that appears as RESULTS a thin layer of water flowing in rhythm with respiration; its visualization excludes PNX with the exception of The feasibility of ultrasound in first time eligible massive lung fibrosis, complete atelectasis and pleural patients was 98.9% (186/188) and 91.2% in all patients symphysis. The comet tail artifacts (or “US B-lines”) admitted with blunt thoracic trauma (186/204 patients). (Lichtenstein 2005) are the well-defined rough vertical The prevalence of PNX revealed by CT scan (gold artifacts arising from the visceral pleura and spreading standard) in the studied population was of 56/186 up to the edge of the screen without fading (Fig. 2). (30.1%), of which 29 were anterior and 27 anterolateral. These artifacts are inconstant findings in nondyspnoic PNX was confirmed by anteroposterior supine chest ra- patients (up to 60%), sometimes present in small num- diograph in 30/56 cases (53.6%), without false positive bers (Ͻ8) in a normal chest and even are often seen in radiographic results (specificity 100%). Despite the ex- injured patients, especially in an expanded lung near a cellent specificity, the plain chest radiographs presented pleural air collection. Comet tail artifacts are absent in false negative results in 26/56 cases (46.4%) (occult PNX because, as they arise from visceral pleura, they are PNX), showing sensitivity of 53.6% (95% CI: 39.8 to hidden in PNX by the air collected in the pleural space. 67.0) and specificity of 100%, a 100% positive predictive We did not apply in our study other accessory US criteria value and a 83.3% (95% CI: 76.6 to 88.8%) negative of PNX diagnosis, such as M-mode and colour-power- predictive value (LR: 70.2). All PNXs missed by stan- Doppler mode, considering them to be superfluous in our dard supine chest radiograph were anterior; 12 of these experience and unjustifiable to delay the examination’s were very small (up to 1 cm thick and seen in no more execution time in emergency. than four contiguous images) (Wolfman et al. 1993). To evaluate the PNX extent, we looked for the Supine chest radiograph correctly identified three PNXs “lung point sign”: this is a dynamic sign that represents classified as anterior by recognition of radiographic deep the fleeting appearance of pleural sliding (with or with- sulcus sign, while it identified all the anterolateral PNXs. out comet tail artifacts) replacing a PNX US pattern Transthoracic ultrasonography detected a PNX in (exclusive horizontal lines of reverberation in place of 55/56 patients (98.2%). The sliding sign was visualized Pneumothorax echo-pattern G G. SOLDATI et al. 1161

Table 1. Signs detected by transthoracic ultrasonography to evaluate the presence and the extent of PNX and to deline the ultrasonographic deep sulcus sign are summarized

Sign Type Definition Anatomical structure Normal PNX

Pleural line Static Horizontal hyperechoic band lying between Lung surface-chest wall interface Yes Yes and below the curved rib lines Bat sign Static Image showing a bat on the wing obtained The pleural line between two Yes Yes by applying the probe longitudinally contiguous ribs over an intercostal space Sliding sign Dynamic Movement of the pleural line synchronised Visceral on parietal pleural layer Yes No with respiration gliding Comet tail Static Vertical hyperechoic artifact arising from Bubbles with gas/fluid interfaces Yes, No pleural line and spreading at the bottom from visceral pleural layer incostant without fading A-line sign Static Exclusive horizontal lines visible at regular Reverberation artifact arising Yes, Yes intervals below the pleural line, typical from the pleural line incostant of the PNX US pattern B-line sign Static Ultrasound finding looked for to rule out Bubbles with gas/fluid interfaces Yes, No PNX and corresponding to the comet tail from visceral pleural layer incostant artifacts Lung point Dynamic Fleeting appearance of sliding sign Reinstatement of contact No Yes, occult between visceral and parietal pleural lyers Deep sulcus sign Static The anterior thoracic area corresponding to Air collecting in antideclive No Yes, occult cardio-mediastinic borders and anterior pleural space costophrenic angles

in all patients without a PNX (130 subjects) while the al. 2001; LoCicero and Mattox 1989). The diagnosis presence alone of comet tail artifacts ruled out PNX in 69 of traumatic PNX is suggested by clinical signs and is patients (53.1%). One occult PNX (very small) was generally confirmed by standard chest radiography. missed by US, but false positive US results were never However, up to 30% of cases are missed (occult) by observed. We found the US sensitivity and specificity in the initial bedside anteroposterior supine chest radio- diagnosing PNX to be 98.2% (95% CI: 90.4 to 99.9%) graph (Tocino et al. 1984). Previous studies of cadaver and 100%, respectively, the positive predictive value have identified the supine position as the least sensi- being 100% and the negative predictive value 99.2% tive in the detection of PNX and shown that readers of (95% CI: 95.8 to 99.8%) (LR 128.7). The negative the supine radiographs were not confident in their predictive value of the comet tail artifacts (B-line sign) diagnosis until 200 up to 400 mL of air had been was 100%. The lung point was recognised in each of the introduced (Rhea et al. 1989). An excellent method for 28 patients with anterior PNX; at this level, comet tail an anteroposterior projection with carryable device artifacts were shown in 25 cases (89.3%). The evaluation requires, in fact, a patient position at least partially of anterior PNX extension by using the lung point agreed upright, an exposition at the end of the inspiration, a enough with CT scan evaluation in all 27 cases, resulting patient/tube distance of 1.5 to 2 m and correct tension in a Ϯ2.3 cm difference (range 1–5 cm), with US over- values (80–100 kV) associate with limited exposition estimation in 14 cases and US underestimation in seven times (Յ0.1 s.) (American College of Radiology cases. Finally, in each US detected anterior PNX, the 1995). location of PNX involved the lower anteromedial zone of On the other hand, the detection of an occult PNX is chest wall as expected, always showing an ultrasono- important, because the progression of small PNX can graphic deep sulcus sign. The average time to complete have a significant deleterious effect on respiratory func- the chest US examination was Յ2 min. tion and the development of tension PNX can occur up to in half of these cases, particularly in the setting of gen- DISCUSSION eral anaesthesia or PEEP (Tocino et al. 1984; Wescott PNX may be the result of preexisting lung dis- and Cole 1983). When a precise recognition is needed, a ease, the complication of diagnostic and therapeutic referral to CT is mandatory (Wolfman et al. 1993). procedures or the outcome of trauma. Chest trauma Despite its excellent diagnostic accuracy, CT has some accounts for 25–33% of all trauma-related deaths and drawbacks, mainly because it involves the removal of PNX represents the second most common injury, after critically-ill patients, resulting in increased delay, costs rib fracture, in blunt chest trauma (Di Bartolomeo et and exposure to irradiation. 1162 Ultrasound in Medicine and Biology Volume 32, Number 8, 2006

Because of its simplicity, security and portability, in the deep sulcus area in normal subjects, were a useful ultrasonography could be especially useful in emergency adjunct to the sliding sign in a subset of patients (53.1%), diagnosis, in which no radiographic equipment is readily because of their absolute negative predictive values. On available. Furthermore, for physical reasons, the detec- the other hand, comet-tail artifacts were more evident as tion of the absence of sliding sign in the supine patient a sign associated with the lung point (89.3%) and repre- with PNX is more successful anteriorly in the parasternal sented a useful finding for its quick recognition. line and near the diaphragm (“deep sulcus sign” area) Our study has several limitations. First, all US ex- (American College of Surgeon 1997). In fact, when the aminations were performed by EPs who are experienced patient is in the supine position, air in the pleural space in US imaging; therefore, we could not assess sonogra- collects anteriorly and inferiorly within the anteromedial pher operator dependence. Second, our studied popula- and costophrenic regions of the pleural spaces. When the tion includes only patients with blunt chest trauma who patient is upright, the air collects in the apicolateral underwent a comprehensive (including CT) chest imag- location. ing. It could be argued that the patients who required a Although it is still a “young” technique in trauma CT, on average, had more serious injuries. Third, the and intensive care, the use of US to detect the presence sample size in this study is relatively small and our of PNX has an established scientific basis. The first inclusion criteria may have introduced bias to it. Fourth, report of ultrasonographic findings about PNX appeared the technique requires further investigation in patients in 1986 in a veterinary journal (Ratanen 1986), while with preexistent disease affecting the pleura. Wernecke et al. (1987) reported the first evaluation of In conclusion, US has to be considered as an elec- PNX by US in humans. The technique has been exam- tive examination for the PNX diagnosis. The extended ined in traumatic, spontaneous and iatrogenic PNX focused assessment with sonography for trauma (Rowan et al. 2002; Sistrom et al. 1996; Targhetta et al. (E-FAST), as recently proposed (Kirkpatrick et al. 2004), 1992; Wescott and Cole 1983). Despite the experience of should be investigated as a clinically attractive modality Sistrom et al. (1996) (sensitivity for PNX 73%, specific- that every physician can readily learn. The finding of an ity 68% and high interobserver variability) and, recently, immobile pleural line in the “deep sulcus sign” area of Kirkpatrick et al. (2004) (sensitivity 58.9% but spec- allows the diagnosis of PNX, whereas the systematic ificity 99.1%), sensitivities greater than 90% have been research of the lung points permits the evaluation of its reported in patients with PNX by other authors (Dul- extent and consequently the hypothesis of its radio- chavsky et al. 2001; Lichtenstein et al. 1999; Targhetta et graphic visibility. Anyway, an US PNX diagnosis, inde- al. 1993). Our previous research (Soldati and Rossi pendently from its clinical or radiographic expression, 2000) and recent unpublished observations confirmed imposes an accurate observation because of the risk of that US is more sensitive than plain radiography and as progression toward more serious conditions. sensitive as CT for PNX detection. The prevalence of occult PNXs seen among all PNXs was 46.5%, higher Acknowledgments—We are indebted to Dr. Monica La Sala for tech- nical assistance during the preparation of manuscript. than recent literature data (Kirkpatrick et al. 2004) and than the ones in our unpublished observations, but it was in agreement with the recent data of Neff et al. (2000). REFERENCES The difference among the various studies around the American College of Radiology. Appropriateness criteria. Reston, VA: sensibility of the US can mainly be imputed to the American College of Radiology, 1995. different experience of the operators and the different American College of Surgeons. Advanced trauma life support course for doctors. Committee on Trauma. Chicago: Instructors Course criteria adopted for the US diagnosis of PNX, while the Manual, 1997. difference in the prevalence of the occult PNXs can Cunnigham J, Kirkpatrick AW, Nicolaou S, et al. Enhanced recognition depend on the different patient inclusion and exclusion of “lung sliding” with power color Doppler imaging in the diagno- sis of pneumothorax. J Trauma 2002;52:769–771. criteria adopted in the single studies. Di Bartolomeo S, Sason G, Nardi G, Scian F, Michelutto V, Lattuada In our study, chest ultrasonography, performed by L. A population-based study on pneumothorax in severely trauma- EPs, focused on evaluating the sliding sign: US was tized patients. J Trauma 2001;51:677–682. Dulchavsky SA, Schwarz KL, Kirkpatrick AW, et al. Prospective more sensitive and accurate than plain radiography per- evaluation of thoracic ultrasound in the detection of pneumothorax. formed by radiologists (98.2 versus 53.5%) and it was J Trauma 2001;50:201–205. nearly as accurate as CT. However, the fact that the Gobien RP, Reines HD, Shabel SI. Localized tension pneumothorax: Unrecognized form of barotrauma in adult respiratory distress radiologists did not know the research project, as well as syndrome. Radiology 1982;142:15–19. they did not have any direct knowledge of the severity of Gordon R. The deep sulcus sign. Radiology 1980;136:23–27. the patients’ condition, could have penalised the estimate Kirkpatrick AW, Sirois M, Laupland KB, et al. Hand held thoracic sonography for detecting post-traumatic pneumothoraces: The ex- of diagnostic accuracy in favour of ultrasonography. tended focused assessment with sonography for trauma (EFAST). Comet-tail artifacts (US B-line sign), inconstant findings J Trauma 2004;57:288–295. Pneumothorax echo-pattern G G. SOLDATI et al. 1163

Lichtenstein D, Menu Y. A bedside ultrasound sign ruling out pneu- with chest radiography and CT—initial experience. Radiology mothorax in the critically ill: Lung sliding. Chest 1995;108:1345– 2002;225:210–214. 1348. Sistrom CL, Reiheld CT, Gay SB, Wallace KK. Detection and estima- Lichtenstein D, Meziere G, Biderman P, Gepner A. The “lung point”: tion of the volume of pneumothorax using real-time sonography. An ultrasound sign specific to pneumothorax. Intensive Care Med Efficacy determined by receiver operating characteristic analysis. 2000;26:1434–1440. AJR 1996;166:317–321. Lichtenstein D, Meziere G, Biderman P, Gepner A. The comet tail Soldati G, Rossi M. Pneumotorace traumatico: Diagnosi ecografica in artifact: An ultrasound sign ruling out pneumothorax. Intensive urgenza. G Ital Ecografia 2000;3:269–274. Care Med 1999;25:383–388. Soldati G. Lung sonography: Artifact, movement or echotexture? G Ital Lichtenstein DA. Pneumothorax and introduction to ultrasound signs in Ecografia 2001;4:329–338. the lung. In: Lichtenstein DA, ed. General ultrasound in the criti- Targhetta R, Burgeois JM, Chavagneux R, Balmes P. Diagnosis of pneumothorax by ultrasound immediately after ultrasonically cally ill. Berlin/Heidelberg, Germany: Springer-Verlag, 2005:105– guided aspiration biopsy. Chest 1992;101:855–856. 115. Targhetta R, Burgeois JM, Chavagneux R, et al. Ultrasonic signs of LoCicero J 3rd, Mattox KL. Epidemiology of chest trauma. Surg Clin pneumothorax: a preliminary work. J Clin Ultrasound 1993;21: North Am 1989;69:15–19. 245–250. Neff MA, Monk JS, Peters K, Nikhilesh A. Detection of occult pneu- Tocino I, Miller MH, Frederick PR, Bahr AL, Thomas F. CT detection mothorax on abdominal computed tomographic scans in trauma of acute pneumothorax in head trauma. AJR 1984;143:989–993. patients. J Trauma 2000;49:281–285. Wernecke K, Galanski M, Peters PE, Hansen J. Pneumothorax: Eval- Ratanen NW. Diagnostic ultrasound: Diseases of the thorax. Vet Clin uation by ultrasound—preliminary results. J Thorax Imaging 1987; North Am 1986;2:49–66. 2:76–78. Rhea JT, Novelline RA, Lawrson J, et al. The frequency and signifi- Wescott J, Cole S. Barotrauma. In: Herman P, ed. Iatrogenic thoracic cance of thoracic injuries detected on abdominal CT scans of complications. New York, NY: Springer-Verlag, 1983:179–209. multiple trauma patients. J Trauma 1989;29:502–505. Wolfman NT, Gilpin JW, Bechtold RE, et al. Occult pneumothorax in Rowan KR, Kirkpatrick AW, Liu D, Forkheim D, Mayo JR, Nicolaou patients with abdominal trauma: CT studies. J Comput Assist S. Traumatic pneumothorax detection with thoracic US: correlation Tomogr 1993;17:56–59.