Methods
Literature review was performed independently by the two authors in MEDLINE (via PubMed)
covering a time span from 2002 to 2017 using free text and Medical Subject Headings/MeSH.
The search terms included “ultrasonography of the lung”, “ultrasound of the lung”,
“ultrasound of pleural disease” and a selected search for “ultrasound of pneumonia”,
“pleural effusion”, “chest ultrasound”, etc. Article selection for the bibliography
was based on consensus according to relevance and evidence. Levels of evidence as
far as mentioned in this article refer to the international evidence-based recommendations
for point-of-care ultrasound published in 2012 [6].
Technical requirements
Depending on the situation and the setting, different conditions and prerequisites
must be met for ultrasound equipment. Suitable transducers for lung ultrasound are
a convex, micro-convex or phased array probe for deeper structure visualization (bandwidth
range 1–6 MHz) and a linear probe for superficial structures and pathology detection,
e. g. pneumothorax (bandwidth range 3–18 MHz) [12]. For quick and easy bedside access, a portable or handheld unit with a micro-convex
probe may be used (evidence level B) [6]. In particular for emergencies, a fast start-up time of the system is advisable.
Preset protocols should be available to obviate the need for time-consuming protocol
adaption on-site. Further criteria are handling issues (compact portable or handheld
unit, easy to clean) [10].
Scanning technique
If possible, the examination should be performed in the upright seated patient (or
at least 45 degrees), allowing for access to the posterior recesses. Otherwise, the
standard scanning position is supine at rest with the patient tilted to the opposite
side of interest, if necessary. Access is acquired through the intercostal spaces.
Limitations are defined by the bones of the chest wall and heart. Furthermore, lesions
located deep inside the lungs covered by layers of healthy aerated tissue are not
accessible. This includes the lung areas adjacent to the paravertebral space as well
as the mediastinum [13].
Different scanning techniques adapted to the clinical setting are advocated in the
literature. A complete lung examination (standard of reference) covers 12 imaging
regions, 6 on each side [10]
[14]. For this, each hemi-thorax is divided into anterior, lateral, and posterior zones
by the anterior and posterior axillary lines. Imaging is performed in the upper and
lower part of each of these regions ([Fig. 1]). In the recumbent critically ill, placing the arm across the chest facilitates
access to the posterior zone [14]. As an extension to this protocol, a 28-site scanning technique has been suggested
for a semi-quantitative assessment of pulmonary congestion over multiple intercostal
spaces [6]
[15].
Fig. 1 Transducer positions for a lung ultrasound examination in the recumbent patient (a: anterior axillary line, b: posterior axillary line, 1–3: anterior, lateral and posterior scanning zones (each
divided into upper and lower part by the dotted line), asterisks: upper and lower
BLUE points (on the left image the transducer is placed on the upper BLUE point),
plus-sign: PLAPS point). A complete examination covers all three zones at each side
scanned in the upper and the lower part [14]. The abbreviated examination covers the two BLUE points and the PLAPS point on both
sides (for the detailed description, how to localize BLUE and PLAPS points see [10]).
Abb. 1 Schallkopfpositionen für die Ultraschalluntersuchung der Lunge am liegenden Patienten
(a: vordere Axillarlinie, b: hintere Axillarlinie, 1–3: vorderer, lateraler und hinterer Untersuchungsbereich
(jeweils durch die gestrichelte Linie in einen kranialen und einen kaudalen Abschnitt
geteilt), Sterne: oberer und unterer BLUE-Punkt (auf dem linken Bild ist der Schallkopf
am oberen BLUE-Punkt aufgesetzt), Plus-Zeichen: PLAPS-Punkt). Die vollständige Untersuchung
deckt auf beiden Seiten des Thorax alle drei Bereiche jeweils im kranialen und kaudalen
Abschnitt ab [14]. Die verkürzte Untersuchung deckt die beiden BLUE-Punkte und den PLAPS-Punkt jeweils
auf beiden Seiten ab (zur Lokalisation der BLUE- und PLAPS-Punkte siehe [10]).
Useful abbreviations of this protocol are recommended for specific clinical questions
and emergency situations with time constraints. The most common approach on international
consensus e. g. for lung interstitial syndrome and pulmonary edema or pneumothorax
omits the posterior scan regions and uses only the anterior and posterolateral chest
wall approaches in supine position, resulting in an 8-zone scanning technique [6]
[10]. A further abbreviated protocol uses only 6 scanning positions, including only two
anterior and one posterolateral position on each side [6]
[10]. In the literature, the two anterior scanning positions have been addressed as BLUE
points (BLUE = bedside lung ultrasound in emergency) and the lateral as the PLAPS
point (posterolateral alveolar and/or pleural syndrome) ([Fig. 1]) [10]. A simplified approach for the quick detection of BLUE and PLAPS points in the emergency
setting using the palm width and the finger length of the sonographer has been described
by Lichtenstein [10]. Any protocols that do not cover the posterolateral scan regions have only limited
indications, e. g. monitoring interstitial lung edema, since they will not allow for
efficient assessment of pleural effusion and posterior atelectasis [6].
Chest wall
In the healthy subject, B-mode lung ultrasound produces clear images of the chest
wall structures (ribs, intercostal space, sub-pleural fat layer) until the marked
change of impedance at the pleural surface. All relevant structures for respiratory
mechanics, including the diaphragm, can be clearly visualized. The motion of the chest
wall and diaphragm can be followed in B-mode and documented in M-mode, e. g. for the
assessment of chronic obstructive lung disease (see below) or phrenic nerve lesions
with diaphragmatic paralysis ([Fig. 2]) [2]
[16].
Fig. 2 Documentation of respiratory motion in M-mode (lung sliding).
Abb. 2 Dokumentation von Atembewegungen im M-Bildmodus (Pleuragleiten).
Normal findings in pleura and lung, terminology
The interpretation of lung ultrasound is based on a set of imaging signs which alone
or in combination are highly sensitive and specific for a number of normal and pathologic
conditions (diagnostic accuracy between 90 % and 100 %). In the healthy subject, a
large proportion of the ultrasound waves is reflected at the pleural line due to the
differences in velocity and acoustic impedance between normal tissue and air-filled
parenchyma (lung, airways, etc.). Hence, the pleural line is marked by a bright signal.
The following space between the pleural signal and the bottom of the image is called
Merlin’s space [10]. It contains a homogeneous, structureless signal from back-scattering of ultrasound
waves by healthy aerated lung tissue. Superimposed on this background, a single line
or multiple lines parallel to the pleural line and perpendicular to the ultrasound
beam can be observed. They are reverberations resulting from multiple reflections
of ultrasound waves between the pleural line and the surface of the transducer [14]. In the previous literature, this artifact was referred to as “A-lines”, but this
term has been omitted on the occasion of the international point-of-care lung ultrasound
consensus publication since it is not specific for any condition (these lines appear
in the regularly aerated lung tissue as well as in pneumothorax) [6], ([Fig. 3a]). However, the absence of horizontal reverberations may help to detect pathology
in the lung (such as pulmonary edema), but it requires dedicated interaction of the
ultrasonographer to optimize their visualization before their absence can be confirmed.
Fig. 3 Normal and physiologic findings on lung US. a Pleural line (transparent arrowhead) and A-lines (filled arrowheads) in the healthy
lung. b To imagine the outline of a flying bat may help to identify the pleural line between
two ribs in difficult imaging conditions (“bat sign”). c B-mode image, the dotted vertical midline indicates the position from which the M-mode
recording in d was obtained. The M-mode image shows the “seashore sign” with speckled appearance
of the aerated lung beyond the pleural line, resembling a sandy beach with waveform
lines above it from the thoracic wall resembling sea waves. Super-imposed short vertical
scatter bands reflect propagation of normal cardiac pulsation at the patient’s heart
rate (“lung pulse sign”, arrows).
Abb. 3 Normale und physiologische Befunde im Ultraschall der Lunge. a Pleuralinie (transparente Pfeilspitze) und A-Linien (weiße Pfeilspitzen) bei gesunder
Lunge. b Bei schwierigen Schallbedingungen kann das Bild einer Fledermaus die Identifikation
der Pleuralinie erleichtern („Fledermauszeichen“). c Die gepunktete Linie im B-Bild markiert die Position, von der das M-Bild in d abgeleitet wurde. Das M-Bild zeigt ein Muster, das an einen Sandstrand (feinkörniger
Aspekt der belüfteten Lunge jenseits der Pleuralinie) und Meereswellen (Reflexionen
der Thoraxwand) erinnert („seashore sign“). Diesem Muster sind pulssynchrone vertikale
Artefakte (Pfeile) überlagert, die den fortgeleiteten Herzbewegungen des Patienten
entsprechen („Lungenpuls“/„lung pulse sign“).
The “bat sign” is a helpful feature for the recognition of the pleural line even in
difficult scanning conditions (patient motion, subcutaneous fat). It can be seen with
the transducer in rectangular position to the ribs. The reflections from the ribs
look like the wings of a bat while the back of its body is outlined by the parietal
pleura [17] ([Fig. 3b]).
For the exclusion of air inside the pleural layers, additional signs such as “lung
sliding” need to be confirmed. Inspiration and expiration result in a lung volume
change and consequently sliding of the visceral pleural surface in relation to the
parietal pleura and chest wall. This movement appears as a homogeneous twinkling (shimmering,
sparkling or glittering) below the pleural line [10]
[14]. The presence of this sign indicates preserved lung ventilation, direct contact
of the visceral pleura to the parietal pleura and therefore the absence of pneumothorax
[18]. A reduced or absent lung sliding sign may indicate hypoventilation, e. g. due to
low tidal volume ventilation, or regional hypoventilation in obstructive lung disease
[14]. However, it may become less visible when the ultrasound beam is tangential rather
than perpendicular to the pleural surface [14]. Another sign of preserved lung physiology is the lung pulse sign, also referred
to as “T-lines” ([Fig. 3c]). It results from relative motion of the pleura due to propagated motion from cardiac
pulsations [14].
For documentation of lung sliding with printed images, M-mode recordings in the healthy
subject show the stationary extrapleural structures and a speckled appearance of the
aerated, moving lung beyond the pleural line. As this image resembles calm waters
(extrapleural structures) washing on the speckled sandy beach (Merlin’s space), it
was given the name “seashore sign” [14] ([Fig. 3 d]).
Pleural disease
As already mentioned, lung US is highly sensitive for any pathology that affects the
marked change of impedance at the pleural surface – i. e. pleural effusion or pneumothorax
[7].
Pleural effusion
With the transducer in rectangular position to the ribs, pleural effusion appears
in a quadrangular space defined by the pleural line (chest wall), the shadows of the
ribs and the lung line (visceral pleura) – also called “quad sign”. If the effusion
contains free, non- organized liquid, the distance between the pleural line and lung
line decreases in inspiration due to the increased volume of the chest and increases
in expiration – “sinusoid sign” [17]. It correlates to the viscosity of the pleural fluid, i. e. for the choice of the
needle size for drainage. In malignant pleural effusion, pleural nodules according
to metastatic growth can be easily detected ([Fig. 4]). In outpatients with acute chest pain, pleural ultrasound suggests the diagnosis
of acute pleurisy when showing pleural thickening and minimal effusion [19].
Fig. 4 Malignant pleural effusion (asterisk) in a 30-year-old male patient with history
of malignant melanoma and progressive shortness of breath. The pleura adjacent to
the diaphragm appears thickened (transparent arrowheads) with adherent nodular structures
related to metastatic growth (filled arrowhead).
Abb. 4 Maligner Pleuraerguss (Stern) bei einem 30 Jahre alten Patienten mit malignem Melanom
und zunehmender Kurzatmigkeit. Die zwerchfellnahe Pleura erscheint verdickt (transparente
Pfeilspitzen) mit anheftenden nodulären Formationen als Korrelat pleuraständiger Filiae
(weiße Pfeilspitzen).
Pneumothorax
Air inside the pleural cavity completely reflects the ultrasound waves at the pleural
line with no available imaging data from the visceral pleura and deeper [3]. Lung sliding/pleural sliding therefore cannot be observed. Its absence has a sensitivity
of 95.3 %, specificity of 91.1 %, and negative predictive value of 100 % (p < 0.001)
for pneumothorax with a low false-positive rate [18]. In M-mode, the absence of motion results in a static pattern of horizontal lines
(“stratosphere sign”) which replaces the seashore sign [17]
[20]. The absence of the lung pulse sign further contributes to the diagnosis [21] ([Fig. 5]).
Fig. 5 Pneumothorax: Presence of horizontal reverberation artifacts, but no lung sliding
can be observed in B-mode imaging (right). The arrowhead indicates a repeat artifact
of the pleural line (so-called “A-line”; per definition corresponding to the distance
of the pleural line to the probe). In M-mode (left), the absence of motion is documented
as a static pattern of horizontal lines (“stratosphere sign”) which replaces the seashore
sign. The absence of the lung pulse further contributes to the diagnosis of pneumothorax.
Abb. 5 Pneumothorax, erkennbar an horizontalen Reverberationsartefakten jenseits der Pleuralinie,
aber ohne Pleuragleiten im B-Bild (rechts). Die weiße Pfeilspitze markiert einen Reverberationsartefakt
der Pleuralinie (sogenannte A-Line, die per definitionem im doppelten Abstand Schallkopf-Pleuralinie
zu beobachten ist). Im M-Bildmodus (links) fehlen die feinkörnigen Reflexe des bewegten
Lungengewebes, indessen zeigen sich die statischen horizontalen Linien des Stratosphären-Zeichens
(„stratosphere sign“). Ein weiterer Hinweis auf das Vorliegen eines Pneumothorax ist
das Fehlen des Lungenpulses.
In the recumbent patient, air inside the pleural space moves anteriorly and the partially
collapsed lung allocates in the dependent part. The scanning position at the lateral
chest with the alternating pattern of pneumothorax into healthy lung is called the
“lung point”. For documentation, M-mode images may be helpful: The static pattern
of the stratosphere sign with parallel horizontal lines (as they result from air inside
the pleural space) changes into the granular “seashore sign” once the lung tissue
comes into contact with the chest wall as an “on/off” pattern. It has an overall sensitivity
of 66 % (75 % in the case of radio-occult pneumothorax alone) and a specificity of
100 % [3]
[22]. In the presence of an air-fluid level (hydro-pneumothorax), the interface between
the pleural effusion (usually anechoic) and the pneumothorax component (absence of
sliding or pulse and absence of B-lines) is called the “hydro point” [21]. The absence of lung sliding but presence of B-lines and/or lung pulse with the
presence of a lung point is diagnostic of septate pneumothorax [21]
In practical application a suggested algorithm for the evaluation for pneumothorax
comprises (1) assessment of the sliding lung sign as the first step (if present, pneumothorax
can be excluded). Then (2), if lung sliding is absent, further evaluation would look
for artifacts (B-lines exclude pneumothorax) and (3) the presence of the lung point
sign (this would confirm pneumothorax). At this step in the algorithm, if the sliding
lung sign, B-lines and a lung point cannot be observed, the presence of a lung pulse
may still rule out pneumothorax while its absence further supports the diagnosis (4)
([Fig. 6]). An overall sensitivity of this algorithm of 92 % for the diagnosis of pneumothorax
has been reported [6]
[14]]. Hence, lung ultrasound is considered more accurate in the diagnosis of pneumothorax
than a bedside supine anterior chest X-ray (evidence level A on international consensus)
and may even replace CT in many cases [6]
[23]. However, an absence of lung sliding, B-lines, lung point and lung pulse will be
similarly observed in conditions such as lung bullae, contusions, pleurodesis or other
causes of adhesion of the pleural leaves and may result in false-positive diagnosis
[6].
Fig. 6 Algorithm for the detection of pneumothorax with lung US (adapted from [6]). caveat: Only the documentation of the lung point sign allows for the sonographic
diagnosis of pneumothorax. An absence of lung sliding, B-lines, lung point and lung
pulse may be also observed in other conditions such as lung bullae, contusions, pleurodesis
or other causes of adhesion of the pleural leaves leading to a false-positive diagnosis
of pneumothorax. Therefore, the absence of the above-mentioned signs only allows for
the diagnosis of “suspected pneumothorax”.
Abb. 6 Algorithmus zur Diagnostik des Pneumothorax mittels Ultraschall (modifiziert nach
[6]). Cave: Nur die Dokumentation des „lung point“-Zeichens sichert die Diagnose ab.
Das Fehlen des Pleuragleitens, fehlende B-Linien und fehlender Lungenpuls ohne Nachweis
eines „lung point“ können auch in anderen Situationen wie Bullae, Lungenkontusion,
Zustand nach Pleurodese oder in anderen Fällen pleuraer Adhäsionen vorliegen. In diesen
Fällen sollte nur der Verdacht auf einen Pneumothorax geäußert werden.
Acute lung disease
Non-aerated lung tissue due to massive lung edema, lobar bronchopneumonia, pulmonary
contusion or atelectasis allows for deep penetration of ultrasound waves. All of these
can be detected readily when adjacent or close to the pleural surface or in the presence
of pleural effusion.
Consolidation and atelectasis
Lung US is highly effective in the detection of consolidation and atelectasis and
differentiation of these from pleural effusion [14]. The collapsed or consolidated lung appears as poorly defined, wedge-shaped, hypoechoic
tissue ([Fig. 7a]). A preserved volume of the lobe due to fluid uptake in consolidation or a volume
loss with dislocation of adjacent structures in atelectasis helps to distinguish between
the two conditions. However, a combination of both patterns is common. Consolidated
lung tissue appears with a tissue-like pattern reminiscent of the liver (“hepatization”)
and with a usually irregular deep boundary to the aerated lung in the case of partial
lobe involvement or regular in the case of whole lobe involvement [24]. This irregularly shaped margin has also been referred to as the “shred sign” [25] ([Fig. 7b]). Hyperechoic small structures within consolidation correspond to “air bronchograms”
([Fig. 7c]) [26]. A positive “dynamic air bronchogram” with centrifugal progression of the air bronchogram
signal in inspiration is suggestive of alveolar consolidation instead of bronchial
occlusion with subsequent lung collapse and atelectasis [25].
Fig. 7 a Pleural effusion (arrowheads) and atelectasis of the laterobasal lower lobe (open
arrowheads). Hyperechoic artifacts delineate the transition from atelectatic to ventilated
lung. b Pneumonia and sub-pleural atelectasis with irregular boundary to the aerated lung
(“shred sign”, arrowheads). c Air bronchogram inside atelectatic lung (open arrowheads).
Abb. 7 a Pleuraerguss (weiße Pfeilspitzen) und Atelektase des laterobasalen Unterlappens (transparente
Pfeilspitzen). Hyperechogene Artefakte demarkieren den Übergang vom atelektatischen
zum belüfteten Lungengewebe. b Pneumonie und subpleurale Atelektasen mit unregelmäßiger Begrenzung zur belüfteten
Lunge („shred sign“, weiße Pfeilspitzen). c Aerobronchiogramm innerhalb der atelektatischen Lunge (transparente Pfeilspitzen).
On radiograms, airway occlusion can be detected only after resorption of gas and atelectatic
collapse of the excluded part of the lung. Lung US already shows an absence of lung
sliding with persisting vibrations from the heart activity at the pleural line (“lung
pulse sign”) while the lung is still inflated (see above, [Fig. 3]). Lichtenstein et al. reported a sensitivity of 93 % and a specificity of 100 %
for this sign for the diagnosis of complete atelectasis after selective intubation
[27]. However, other conditions associated with alterations in lung sliding need to be
taken into account (e. g., lung fibrosis, pleural adhesion, diaphragmatic dysfunction
and increased abdominal pressure). If a “lung pulse” is already present before intubation,
ultrasound cannot predict atelectasis after selective intubation [26]
[27].
Besides X-ray, ultrasound imaging serves as a fast and easily available bedside test
for the assessment of lung aeration during mechanical ventilation [28]
[29]. Alternatively, only electrobioimpedancy (EIT) provides real-time monitoring of
chest impedance correlating with air content or fluid accumulation, but unlike X-ray
and ultrasound, ventilation maps of the lung generated with EIT do not contribute
information about the morphology of the affected tissue [30].
B-lines/B-pattern
The further investigation of diffuse or focal parenchymal disease is based on hyperechoic
vertical artifacts, summarized as “B-lines” or “B-pattern” [31]. They arise from the pleural line and indicate thickening of interlobular septa
due to congestive or fibrotic changes. These hyperechoic artifacts are long, relatively
sharply delineated and obliterate horizontal reverberations from the pleural line
([Fig. 8]). Due to their origin from the pleural line, B-lines follow lung sliding during
respiration. Instead of more descriptive terms, such as comet tail or rocket sign
(for single and multiple B-lines), it is recommended to use the generic term “B-pattern”
[6]. Per definition, the appearance of B-lines rules out pneumothorax since they could
not be generated through an air filling of the pleural space [32].
Fig. 8 Single a and multiple b B-lines (vertical bright lines) in ICU patient with lung interstitial edema. c confluent B-lines (“white lung”) in a patient with ARDS.
Abb. 8 Einzelne a und multiple b B-Linien (helle, vertikale Linien) bei einem intensivmedizinisch versorgten Patienten
mit interstitiellem Lungenödem. c konfluente B-Linien („weiße Lunge“) bei einem Patienten mit ARDS.
Notably, single B-lines are not necessarily a sign of critical illness, since they
may be found in healthy, elderly subjects (in one study they were present in 37 of
100 persons aged more than 65 years, with 27/37 having 3 or less lines per field of
view). Horizontal reverberations from the pleural line (formerly called “A-lines”)
instead become rare with age (present in only 6/100) [33]. In the newborn, a decrease of B-lines on ultrasound and an increase of lung compliance
during the first 24 h after birth reflect the clearance of lung liquid [34].
Septal B-lines indicate edematous or fibrous thickening of interlobular septa. They
typically appear at a distance of 6 or 7 mm apart according to the anatomic distance
between two interlobular septa ([Fig. 8]). Sets of at least three hyperechoic B-lines arising from the pleural line in one
intercostal space are indicative of interstitial lung syndrome. In dyspneic patients,
they are indicative of diffuse alveolar-interstitial involvement and allow for fast
bedside diagnosis of pulmonary congestion or interstitial pneumonia [35]
[36]
[37]. With this, lung ultrasound for B-lines is superior to supine anterior chest X-ray
in the detection or the exclusion of significant interstitial syndrome (evidence level
B) [6]. In acute dyspnea, the presence of B-lines is a useful discriminator between pulmonary
edema (B-lines present in 100 %) and a respiratory cause, e. g. exacerbation of COPD
(B-lines absent in 92 %) [38]
[39].
As B-lines may be influenced by multiple factors, ongoing investigation is done to
further determine the role of different transducers and the clinical significance
of findings on higher resolution ultrasound imaging [31]. Li et al., for example, investigated the diagnostic value of pleural line changes
in conjunction with B-lines. The coexistence of B-lines with more than one abnormal
pleural line was found in about 30 % of cases: 1) a slightly rough pleural line with
confluent B-lines corresponded to ground-glass opacity on CT; 2) an irregular and interrupted
pleural line with confluent B-lines – parenchymal infiltration; 3) a fringed pleural
line and confluent B-lines – superimposed ground-glass and irregular reticular opacities;
4) a fringed pleural line with scattered B-lines – irregularly thickened interlobular
septa; 5) a wavy pleural line indicated subpleural emphysema [40].
Differential diagnosis with respect to other vertical (formerly called “comet tail”)
artifacts is based on the origin. “E-lines” originate from emphysema in the chest
wall and thus do not follow lung sliding. “Z-lines” appear as vertical hyperechoic
artifacts arising from the pleural line, but do not reach the distal edge of the screen,
typically found in healthy persons and patients with pneumothorax [17]. “T-lines”, as they result from the lung pulse sign, have already been mentioned
above.
Interstitial lung edema
In practical application, a fast anterior two-region scan allows confirmation or ruling
out of acute lung edema with high diagnostic accuracy [7]. Notably, the change from A- to B-pattern in lung congestion appears at a threshold
of 18 mmHg of pulmonary artery occlusion pressure and serves as a direct biomarker
of clinical volemia [17]. In interstitial edema, the number of B-lines correlates with fluid accumulation
and the extent of parenchymal changes on CT [41]. When originating from accumulation of intra-alveolar fluid (ground glass opacities
on X-ray or CT), they appear in an even higher number and are allocated closer to
each other compared to septal B-lines (3 mm or less apart correlating with the degree
of reduced aeration). For a semi-quantitative assessment of the severity of pulmonary
congestion, it has been suggested to count the number of B-lines in multiple intercostal
spaces (28-site scanning technique) or the number of positive scans (eight-region
technique) [6]
[14].
This knowledge can be used for fluid administration management in shock patients (fluid
administration limited by lung sonography = FALLS protocol). After exclusion of pericardial
effusion/tamponade, acute pulmonary embolism (enlargement of the right ventricle),
pneumothorax (A-pattern) and pulmonary congestion (B-pattern) with lung US, fluid
administration is started until the appearance of B-lines indicates subclinical iatrogenic
interstitial edema [17]. A lack of clinical improvement following this fluid administration is considered
diagnostic for distributive septic shock [17].
Lung disease with inhomogeneous distribution
The differential diagnoses for an interstitial syndrome pattern on lung ultrasound
include pulmonary edema of various causes including ARDS, interstitial pneumonia or
pneumonitis or diffuse parenchymal lung disease (lung fibrosis). A focal (localized)
sonographic B-pattern of interstitial syndrome is also found in some cases of atelectasis,
pulmonary contusion, pulmonary infarction and neoplasia. The differentiation is based
on the distinction of focal and diffuse sonographic pattern and other sonographic
signs [6].
ARDS
Non-cardiogenic pulmonary edema related to ARDS is distinguished from cardiogenic
edema by the more inhomogeneous distribution of B-lines and inhomogeneous involvement
of lung parenchyma with “spared areas” of normal parenchyma ([Fig. 9]). Together with the absence or reduction of lung sliding and the presence of pleural
line abnormalities (irregularly thickened and fragmented appearance of the pleural
line), ARDS can be clearly differentiated (evidence level B) [6]
[42]. Further signs suggesting the presence of ARDS are consolidations, pleural effusion
and a lung pulse sign [42]. Similarly, respiratory distress syndrome in neonates presents with bilateral, confluent
B-lines, absence of spared areas and pleural line abnormalities, making it a highly
sensitive test for this condition in the pediatric setting and obviating the need
for chest X-ray [6]
[43]. In suspected transient tachypnea of the newborn (TTN), the diagnosis is established
by the finding of bilateral confluent B-lines in the dependent areas of the lung (“white
lung”’) with a normal or near-normal appearance of the lung in the superior fields
(see above, [Fig. 8c]) [6]
[44]. In pregnant women lung US can serve to observe pulmonary edema in preeclampsia,
to avoid intravenous or excess fluids, which can lead to respiratory failure [45].
Fig. 9 “Spared areas” of intact lung (arrowheads) adjacent to “white lung” (open arrowheads)
in a patient with ARDS.
Abb. 9 Verschonte Areale intakter Lunge („spared areas“, weiße Pfeilspitzen) neben „weißer
Lunge“ (transparente Pfeilspitzen) bei einem Patienten mit ARDS.
Pneumonia
Reissig et al. conducted a prospective, multicenter study on the accuracy of lung
ultrasound in the diagnosis and follow-up of community-acquired pneumonia [9] ([Fig. 10]). In 362 patients in 14 centers, lung ultrasound had a sensitivity of 93.4 % and
a specificity of 97.7 % for pneumonia compared to the reference of final clinical
diagnosis. Breath-dependent motion of infiltrates was seen in 97.6 % of the cases,
an air bronchogram in 86.7 %, blurred margins in 76.5 % and a basal pleural effusion
in 54.4 % [9]. Bourcier et al. have discussed that lung US might even replace routine chest radiography
in the first-line diagnosis of acute community-acquired pneumonia and to reserve chest
CT scan for complicated cases [15]. On the other hand, 8 % of the pneumonic lesions were missed in the study of Reissig
et al. Therefore, they concluded that a normal lung ultrasound examination does not
reliably exclude pneumonia [9]. These numbers appear in a different light when compared to chest radiograms in
recumbent patients, which are frequently of poor quality and difficult to interpret.
With lung US, the examining physician can adjust to the individual situation and repeat
scanning without having to be concerned about radiation exposure. Only recently, Alzahrani
et al. have provided a systematic review and meta-analysis on the diagnostic accuracy
of ultrasound for the diagnosis of pneumonia versus radiological imaging with X-ray
and CT (including 20 studies with an total of 2513 subjects). The pooled sensitivity
and specificity for the diagnosis of pneumonia by lung ultrasound were 0.85 (0.84–0.87)
and 0.93 (0.92–0.95), respectively. The overall pooled positive and negative LRs were
11.05 (3.76–32.50) and 0.08 (0.04–0.15) and the area under the pooled ROC (AUC for
SROC) was 0.978 [46].
Fig. 10 Lateral left lower lobe subpleural pneumonia in a 61-year-old female (M. pneumoniae);
left: MSCT coronal reformation; middle: transverse CT slice; right: ultrasound shows
hypoechoic irregular lung infiltration (arrowheads) through the left lateral 8th intercostal
space.
Abb. 10 Links laterobasales subpleurales pneumonisches Infiltrat, 61 Jahre alte Patientin
(M. pneumoniae); links: koronare CT-Rekonstruktion; Mitte: transversale CT-Schicht;
rechts: Die Sonographie in Höhe des 8. links lateralen Zwischenwirbelraumes zeigt
eine hypoechogene, unregelmäßige Infiltration der Lunge (Pfeilspitzen).
In pregnant patients, community-acquired pneumonia is the most common fatal non-obstetric
infectious complication and a common cause of hospitalization [47]. Diagnosing pneumonia in pregnancy by lung US can affect pregnant women by determining
health risks for the mother and fetus, while avoiding unnecessary X-ray exposure.
However, further studies are required to validate the shift of imaging modality on
a larger scale [48] ([Fig. 11]).
Fig. 11 Anterior right upper lobe pneumonia in a 33-year-old pregnant woman (left: PA chest
X-ray, right: Lung ultrasound through the right anterolateral 3 rd intercostal space).
The ultrasound examination shows only a small lesion (arrowheads) due to limited ultrasound
penetration through well aerated tissue surrounding the consolidation.
Abb. 11 Pneumonisches Infiltrat im anterioren rechten Oberlappen einer 33 Jahre alten, schwangeren
Patientin (links: Röntgenübersicht p.-a., rechts: Ultraschall durch den dritten rechts
anterolateralen Interkostalraum). Die Sonographie zeigt nur eine kleine pleuraständige
Läsion (Pfeilspitzen), da interponiertes, belüftetes Lungengewebe die tiefer gelegenen
Anteile des Infiltrates maskiert.
Similar ultrasound signs of pleural diseases and pneumonia are found in pediatric
patients. For the diagnosis of pneumonia in children, lung ultrasound is considered
as accurate as chest radiography (international consensus, evidence level A) [6]
[49]. According to this guideline, a positive lung ultrasound in children with suspected
pneumonia obviates the need for additional radiography. In contrast to adult patients,
intravenously injected ultrasound contrast media (e. g. SonoVue, Bracco, Milan) already
play a role in imaging and intervention in complex pneumonia in children. They have
been applied successfully to distinguish between necrotic and vital lung in necrotizing
pneumonia when grayscale US was not conclusive. Atelectasis can be readily differentiated
from empyema by the vascular enhancement ([Fig. 12]). For interventions, this may be helpful to establish the indication for catheter
placement and selection of the proper position for drains [50]. Intracavitary contrast material application can help to establish the indication
for drainage, detection of septa hindering the complete drainage of the pleural effusion
(e. g. in multi-septated effusion suggesting the use of Streptokinase) and to finally
choose the appropriate timing for catheter removal [44].
Fig. 12 Pleuropneumonia. 68-year-old male patient, symptomatic 4 days with febrile temperature,
productive cough, left lower chest pain, elevated inflammatory markers in laboratory
tests (CRP > 200 mg/L). On X-ray left lower lobe posterior segmental pneumonia (B – arrowhead) and obscured costodiafragmal sinus due to effusion (A – arrowhead). The lung ultrasound shows local consolidation with few air bronchograms
(bright spots on the dark grey background), (C–arrows – infiltration, arrowhead – air bronchogramms) on the left lower lung lateral
surface with adjacent black pleural effusion (D – arrow) and marked hypervascularization in the lesion (E – color in color Doppler mode, F – microvascularization mode), suggestive of pleuro-pneumonia.
Abb. 12 Pleuropneumonie: 68 Jahre alter Patient, seit 4 Tagen symptomatisch mit Fieber, produktivem
Husten, links basalem Thoraxschmerz und erhöhten Entzündungswerten (CRP > 200 mg/l).
Die Röntgenübersicht zeigt ein segmentales Infiltrat des posterioren linken Unterlappens
(B – Pfeilspitze) und einen Randwinkelerguss (A – Pfeilspitze). Der Lungen-Ultraschall zeigt links laterobasal lokale Konsolidationen
mit einigen wenigen Aerobronchogrammen (C – weiße Flecken vor dem dunklen Hintergrund des atelektatischen Lungengewebes; Pfeile
– Infiltration, Pfeilspitze – Aerobronchiogramm), angrenzend Pleuraerguss (D – Pfeil). Markante Hypervaskularisation des Befundes (E – Farbdopplermodus, F – Powerdoppler-Modus).
Chronic lung disease
Pulmonary fibrosis
Given the high sensitivity of lung ultrasound regarding changes in the peripheral
lung, its capacity to detect interstitial lung disease related to pulmonary fibrosis
is not surprising [51]. Interstitial fibrosis with thickened interlobular and intralobular septa results
in an inhomogeneous, diffuse B-line pattern. The combination with pleural abnormalities
and sub-pleural abnormalities that correlate with the typical findings on CT in these
conditions helps to differentiate from cardiogenic pulmonary edema [6]
[52]. In patients with systemic sclerosis, a high sensitivity of lung ultrasound for
early findings of lung and pleura involvement before the onset of clinical symptoms
has been shown and it has been suggested to use it as a routine test to identify patients
for selective referral to HRCT [53]
[54]. A comprehensive assessment, e. g. for scientific use, includes evaluation at 50
positions in intercostal spaces, but for practical application, an abbreviated protocol
would include 14 positions (the second intercostal space parasternal, the fourth along
the mid-clavicular, anterior axillary and mid-axillary lines, and the eighth along
the paravertebral, sub-scapular and posterior axillary lines) and can be performed
in less than 10 minutes [55]. Instead of this time-consuming manual assessment of B-line frequency, software-based
analysis has been suggested to classify the degree of pulmonary edema and pulmonary
fibrosis with high feasibility and over different levels of severity [52]
[56].
Obstructive lung diseases
Generally, the hyperinflation of lung tissue due to emphysema increases the artifacts
and makes lung ultrasound even more difficult. Consequently, nonspecific signs of
air dominance beneath the pleura such as horizontal reverberations from the pleural
line (formerly called “A-lines”) and other lines parallel to these are the predominant
pattern [17]. Functional impairment of respiratory mechanics can be estimated from diaphragm
muscle thickness and diaphragmatic motion. Patients with high risk prior to general
anesthesia, mechanical ventilation, or invasive procedures can be identified [57]. While the observation of diaphragmatic motion with B-mode imaging is the key to
diagnosis, M-mode recordings can be used to document the displacement. In analogy
to the time/volume curve of spirometry, the calculated M-mode Index of Obstruction
(MIO) from the ratio between forced diaphragmatic excursion in the first second and
the maximal expiratory diaphragmatic excursion correlates linearly with airway obstruction
[58]. In the general clinical setting, the detection of abnormalities in respiratory
motion can contribute to the early detection of COPD as an ancillary finding in abdominal
ultrasound [59].
Pulmonary malignancy
Pulmonary involvement is common in various oncologic diseases. While endobronchial
or esophageal ultrasound have become an important modality for the evaluation of mediastinal
lymph node metastases [60], peripherally located focal primary tumor (rarely) or secondary metastatic lesions
(frequently) may still be depicted on transthoracic ultrasound if they are located
subpleural or invade thoracic wall structures [61]. The hallmark of these lesions includes the hypoechoic lesion or multiple lesions
intrapulmonary adjacent to the pleural space or thoracic wall, with marked blood supply
shown on color Doppler, microvascular imaging and contrast-enhanced ultrasound (fast
hemodynamics of contrast media with wash-in and fast wash-out), which helps with respect
to differential diagnosis from pulmonary embolism (avascular lesion) [62]. However, a differentiation of malignancy and focal lesions in pneumonia may be
difficult [63]. Therefore, these features are appreciated for planning and performing US-guided
biopsy, i. e. to detect the best vascularized portion and to obtain representative,
viable tissue samples [64].
Pulmonary vascular disease
The clinical workhorse in the diagnosis of acute pulmonary embolism (PE) is computed
tomography angiography (CTA). However, a fast and reasonably sensitive bedside test
for this acute, life-threatening condition would be desirable on many occasions. Lung
ultrasound is blind to central pulmonary embolism, but obviously a substantial number
of pulmonary emboli involve the peripheral lung. The ultrasonography hallmark of acute
PE in the periphery is triangular, hypoechoic, and pleural-based parenchymal lesions
due to peripheral pulmonary infarction with localized and/or basal effusion (from
radiography also known as Hampton’s hump) [65] and abrupted vessel on color Doppler imaging corresponding to a hypovascular area
[66]. In a prospective multicenter trial based on these signs, Mathis et al. demonstrated
a sensitivity of 74 % and a specificity of 95 % for lung ultrasound for the diagnosis
of peripheral pulmonary embolism (352 patients with a PE prevalence of 55 % confirmed
by CTPA, positive predictive value 95 %, negative predictive value of 75 %, accuracy
of 84 %) [66]. The diagnostic confidence was highest in the presence of two or more typical triangular
or rounded pleural-based lesions. PE was considered probable if there was one typical
lesion in combination with pleural effusion. Small (< 5 mm) subpleural lesions or
a single pleural effusion alone was considered a sign for possible PE. On international
consensus level lung ultrasound has been acknowledged as an alternative diagnostic
tool for the diagnosis of PE when CT is contraindicated or unavailable (evidence level
A) [6]. Ongoing research is focused on contrast-enhanced ultrasound with intravascular
contrast agents for the diagnosis of perfusion deficits in acute pulmonary embolism,
but its role in the routine is not yet defined [67].
The full picture: Comprehensive assessment of pulmonary disease
The fast and efficient application of the above-described knowledge for the diagnosis
of acute respiratory failure defines the need for a structured approach. For this
purpose, Lichtenstein et al. have advocated the “bedside lung ultrasound in emergency
(BLUE) protocol” [7]. It covers important pathologies such as pneumonia, congestive heart failure, COPD,
asthma, pulmonary embolism and pneumothorax. It can be performed in less than 3 minutes
within the first 20 minutes of admission and would be completed with echocardiography
[17]. The first step in the algorithm is the detection of lung sliding, then the predominant
artifact pattern is assessed. Lung sliding with predominantly horizontal reverberations
(“A-pattern”) is suggestive for asthma or COPD. Lung sliding with bilateral appearance
of an anterior B-pattern indicates pulmonary edema. Lung sliding with a normal anterior
appearance of the lung plus deep venous thrombosis is suggestive for pulmonary embolism.
Absent lung sliding with horizontal reverberation artifacts (“A-pattern”) and with
a lung point sign is found in pneumothorax. Independently on lung sliding, anterior
alveolar consolidations, anterior diffuse B-lines, anterior asymmetric interstitial
patterns, posterior consolidations or effusions are found in different types of pneumonia.
The sensitivity and specificity of these findings range from 80–100 %. Overall a correct
diagnosis was established in 90.5 % of cases [7]
[17].
