DOI:10.2214/AJR.04.1672
AJR 2006; 186:58-66
© American Roentgen Ray Society
The Role of Imaging During Extracorporeal Membrane Oxygenation in Pediatric Respiratory Failure
Alex M. Barnacle1,
Liz C. Smith2 and
Melanie P. Hiorns1
1 Department of Radiology, Great Ormond Street Hospital for Children, Great
Ormond St., London WC1N 3JH, England.
2 ECMO Coordinator, Great Ormond Street Hospital for Children, London WC1N 3JH,
England.
Received October 27, 2004;
accepted after revision June 13, 2005.
Address correspondence to A. M. Barnacle.
Abstract
OBJECTIVE. Extracorporeal membrane oxygenation (ECMO) is
increasingly widely used in pediatric respiratory failure. Despite playing a
key part in patient management during ECMO, the role of radiology is not
widely reported. We discuss the principles of ECMO support and the normal
imaging appearances. Radiologic findings arising from the complications of
ECMO are highlighted.
CONCLUSION. Radiology has a central role in establishing
well-designed imaging protocols and vigilant reporting of ECMO apparatus
positions and complications.
Keywords: catheters • chest • ECMO • neonates • pediatric radiology • respiratory failure
Introduction
The use of extracorporeal membrane oxygenation (ECMO) as a means of
therapeutic support in infants and children with severe respiratory disease
has expanded dramatically over the past two decades. Imaging plays a key role
in the management of patients both before and during ECMO support. It is vital
that pediatric radiologists in ECMO centers apply well-designed imaging
protocols and are vigilant in the reporting of apparatus positions and
complications of treatment.
The role of radiology in modern ECMO support is not widely reported. Our
article discusses the normal radiologic appearances during ECMO. The range of
bypass cannula positions is illustrated, and appropriate imaging protocols are
discussed. The radiologic findings arising from the complications of ECMO are
highlighted.
Background
ECMO is a modified pulmonary or cardiopulmonary bypass technique used to
support patients with severe cardiac or respiratory failure (or both)
unresponsive to conventional ventilatory support and medical treatment. The
standard indications for the use of ECMO remain neonatal conditions such as
meconium aspiration, primary pulmonary hypertension, and congenital
diaphragmatic hernia, and it is in the neonatal population that ECMO has been
proven effective as a therapeutic intervention
[1]. More recently, the
indications for ECMO have broadened as techniques and survival rates have
improved, with ECMO increasingly used as a temporary bridge to definitive
treatment in older children with cardiac disease and in patients with
overwhelming sepsis and multiorgan failure
[2].
ECMO involves bypass of venous blood to an external membrane oxygenator
before reintroduction either to the arterial circulation, providing heart and
lung support, or to the venous circulation, providing only lung support. This
allows reduction of barotrauma to the lung parenchyma, providing rest and
recovery. Standard ECMO circuits use a venous-to-arterial circulation, termed
"VA ECMO," whereas more recently the advent of double-lumen venous
cannulae has encouraged the widespread use of venovenous circuits, termed
"VV ECMO." Venovenous ECMO systems deliver oxygenated blood to the
right side of the heart and therefore do not provide mechanical circulatory
support, relying on left ventricular function for delivery of oxygenated blood
to the systemic circulation. The circuit is highly dependent on optimal venous
drainage to maximize the proportion of the circulating blood volume that can
be oxygenated [3]. This method
does, however, offer significant advantages over venous-to-arterial ECMO:
avoiding ligation of the carotid artery, normalizing both preload and
afterload on both ventricles, maintaining pulsatile blood flow, and allowing
perfusion of well-oxygenated blood to the pulmonary circulation and coronary
arteries [3]. Some evidence
suggests that there is an increased risk of neurologic complications with
venous-to-arterial ECMO compared with venovenous ECMO
[2,
4-6],
whereas other studies have failed to confirm such findings
[7]. Some practitioners
recommend the use of a cephalad venous catheter to improve venous outflow from
the cerebral vasculature
[8].
Selection criteria for consideration of ECMO may vary among institutions.
The presence of potentially reversible cardiac or respiratory failure is
central in determining those who will benefit from ECMO support. Parameters,
such as the oxygenation index (OI), serve to provide evidence of severe
respiratory failure despite maximal medical management. The OI is a calculated
index of respiratory function using the values of inspired oxygen
(FIO2 [%]) and arterial oxygen
(PaO2[mm Hg]) concentrations and mean airway pressure
(MAP [cm H2O]) to estimate oxygen exchange and is expressed by the
following formula:
 |
Patient selection is based on patient weight and gestational age,
underlying diagnosis, and absence of complications such as significant
intracranial hemorrhage because heparinization of the ECMO circuit will only
extend any preexisting intracranial bleed. Strict imaging protocols play an
important role in identifying those patients with adverse intracranial
features before commencement of ECMO. Pretreatment cranial sonograms are
obtained in all infants with an open fontanelle. All centers exclude potential
ECMO candidates with evidence of intracranial hemorrhage greater than grade 1.
Attempts have been made to predict the risk of adverse neurologic events on
the basis of pre-ECMO cranial sonography; it appears that infants with severe
cerebral edema or periventricular leukomalacia before ECMO are at greater risk
of subsequent major intracranial complications
[9].
During the period of ECMO therapy itself, imaging has a limited role.
Radiologic investigations serve to identify and monitor complications, a
proportion of which may be clinically unsuspected
[10], and imaging findings may
be contributory in the decision to withdraw treatment after a prolonged course
of ECMO. Recovery, however, is best assessed clinically by improvements in
lung compliance and gas exchange. Levels of support are likely to be adjusted
after trial weaning of ECMO therapy rather than after radiographic progress.
Improvements in the appearance of the chest radiograph are often not expected
until late in the course of ECMO therapy and may lag behind clinical
recovery.
ECMO Cannulae: Type and Position
There is a wide range of pediatric ECMO cannulae in use, each with a
different radiographic appearance. Familiarity of the radiologist with the
type of cannulae used in each institution is essential.
Both arterial and venous cannulae are usually inserted via a surgical
cutdown technique, during which the cannula is sutured both to the vessel wall
and to the skin, and the vessel is tied off. After removal of the cannula,
reconstruction of the vessel may be attempted, depending on the method of
insertion and the duration of therapy. A percutaneous insertion method without
imaging guidance has been described for venous cannulae, as has an
open-assisted technique [11,
12]. This modified Seldinger
technique obviates ligation of the jugular vein and allows flow of blood past
the cannula, reducing potential impairment of cerebral venous drainage. Such a
technique aims to avoid risks to the structures of the neck inherent in the
non-imaging-guided percutaneous procedure.
Desaturated blood is optimally drained from the right atrium through a
venous cannula placed via the right internal jugular vein. The venous cannula
has both end and side holes to enhance drainage. The cannula tip should lie
approximately at the level of the eighth-ninth ribs posteriorly (the expected
level of the right atrium), thus ensuring that both tip and side holes lie
within the atrium. A number of venous cannulae have a radiolucent distal
segment, and the tip may therefore lie more distally than suspected on
radiographic imaging; the reporting radiologist should be familiar with the
cannula type in use. Some types have a radiopaque tip beyond the radiolucent
segment (Fig. 1). This tip can
be difficult to visualize; a conservatively collimated radiograph may fail to
show the tip of a malpositioned line within the inferior vena cava. In
neonates with small-caliber vessels, a venous cannula that is placed too
distally will result in both obstruction to peripheral venous return and
blockage of cannula side holes, leading to inadequate drainage. If the cannula
tip is placed too proximally, there is a risk of the cannula side holes lying
outside the vessel lumen or of dislodgement of the cannula itself, causing
life-threatening air emboli and potential hemorrhage.
Alternative venous cannulation sites include the left internal jugular vein
and the femoral veins. In large children, a second venous cannula is required
to allow adequate exchange of the patient's circulating volume. In such cases,
two venous cannulae are placed simultaneously, usually within the jugular and
femoral veins. The line tips must be placed a sufficient distance from each
other to reduce recirculation of oxygenated blood
(Fig. 2).
Arterial cannulae are usually inserted via the right common carotid artery
with the tip lying within the innominate artery. The tip of the cannula has a
single end hole and is typically radiopaque; it should lie at the origin of
the common carotid artery to maximize delivery of oxygenated blood to the
aortic circulation without causing obstruction to flow within the aortic
lumen. In cases in which dense lung parenchymal opacification obscures the
cardiovascular landmarks, the cannula tip should therefore lie at the level of
the second–third ribs posteriorly because this position correlates with
the origin of the common carotid artery. The exact position of the cannula tip
is difficult to determine on chest radiographs and is confirmed by
echocardiography after placement. Failure to adequately advance the arterial
catheter within the carotid artery increases the risk of inadvertent
dislodgement (Fig. 3). Equally,
cannulae that are placed too far into the vessel may obstruct aortic flow and
increase afterload on the left ventricle. Note that, as with various venous
catheters, some types of arterial cannulae also have a distal radiolucent
segment with a radiopaque tip to confirm the distal position of the cannula
(Fig. 4).
View larger version (174K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —Chest radiograph of 29-month-old female illustrates incorrect
placement of arterial cannula (arrow), which is too high and resulted
in inadvertent displacement of cannula with subsequent life-threatening
hemorrhage.
|
|
View larger version (166K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Chest radiograph of 1-month-old male shows arterial cannula
has distal radiolucent segment. Tip of cannula is gauged by position of
radiopaque tip (arrow), which lies approximately at level of aortic
arch. Echocardiography (not shown) would confirm position of cannula. There is
also kink in upper venous catheter.
|
|
Venovenous ECMO, via a single cannula with a double lumen, is increasingly
popular, with 12- to 18-French cannulae now available allowing support in
patients with a body weight of up to 15 kg. The small-caliber lumens possible
within a single catheter limit the volume of venous exchange possible;
therefore, use of double-lumen venovenous ECMO is restricted by the size of
the patient. The tip of the venovenous cannula should lie within the right
atrium with the smaller arterial lumen directed toward the tricuspid valve to
maximize delivery of oxygenated blood during systole; orientation of the
catheter can be determined on chest radiography or echocardiography
(Fig. 5).
Suturing the cannula within the vessel and patient positioning can cause
inadvertent kinking or narrowing of the catheter, leading to suboptimal
circuit drainage (Fig. 6). This
is particularly common in soft-walled venovenous cannulae, many of which are
made of a material with inherent "memory" so that kinked cannulae
cannot be salvaged and may have to be replaced.
Normal Radiographic Appearances During Therapy
Generalized lung opacification typically occurs when patients in
respiratory failure are commenced on ECMO. This phenomenon does not correlate
with severity of lung disease, but is likely to reflect changes in pulmonary
hemodynamics and physiology and the abrupt decrease in airway pressure
[13-15].
Dense opacification of the lung parenchyma hinders the assessment of the
position of support apparatus, with loss of normal mediastinal landmarks;
hence, bone landmarks (rib and vertebral body levels) become more important.
Particular care should be taken to correctly analyze the positions of ECMO
cannulae and other catheters, recognizing that poor patient positioning and
differing radiographic projections can significantly alter projected line tip
positions. The patient is usually ventilated via an endotracheal tube on
minimal airway pressures, in air, with a rate of approximately 10 breaths per
minute. Alternatively, continuous positive airway pressure techniques may be
used in patients with significant air leaks or multiple chest drains in situ.
Regular chest physiotherapy and progressive tissue healing lead to a gradual
improvement in lung aeration over time, monitored via serial chest
radiographs.
View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 10 —Cranial sonographic image of 4-day-old male obtained using
high-frequency linear probe shows right-sided echogenic subdural hemorrhage
abutting falx (arrow). Extraaxial space (between calipers)
on left side remains typically widened.
|
|
Currently, CT is rarely used in the evaluation of the chest during
treatment. Imaging of the lung parenchyma during the first 7-14 days of ECMO
support is of limited value, given that a period of normal lung recovery must
be allowed before treatment withdrawal can be considered. Recent work suggests
that cross-sectional imaging may have a role in complex cases, particularly in
patients with an unexplained delay in clinical improvement
[16,
17]. Such imaging requires
transportation of the patient to the radiology department, thereby increasing
the risk of complications such as cannula dislodgement. Contrast medium must
be administered via the arterial side of the circuit, distal to the membrane
oxygenator, or directly into the arterial circulation to prevent dilution of
contrast medium in the ECMO system
[16].
View larger version (156K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 11B —Chest radiographs of two infants with pneumothorax.
12-month-old female. Image shows mediastinal free air and left-sided
pneumothorax, with thymus gland (arrows) outlined by mediastinal
air.
|
|
Abdominal radiographs may be requested to show the position of femoral
cannulae. Radiographs show relatively little air within the abdomen during
ECMO. This is a normal finding during treatment due in part to sedation of the
patient with reduced or absent air swallowing. If available, abdominal
radiographs should be closely evaluated for the presence of necrotizing
enterocolitis during support of preterm infants.
The Role of Sonography
Sonography is often the most useful imaging investigation of both the chest
and the abdomen once lung parenchymal opacification has occurred and the
volume of bowel gas within the abdomen becomes limited. Sonography is readily
available within the intensive care setting, allowing minimal disturbance of
the patient or support apparatus. Fluid collections are well depicted, and
there may be clues as to the complexity of the collection: The sonographic
appearances may include fluid collections that are anechoic, septated,
hyperechoic, mixed, or complex. Some studies have shown that collections that
are anechoic or only have a few simple septa are less likely to be infected,
and therefore just represent transudates, than those that show increased
echogenicity or a complex pattern of septations
[18], which may indicate
infection. Other series have shown that neither sonography nor CT can reliably
identify the stage of pleural infection
[19] and that therefore
sonography findings need to be interpreted in their clinical context. Acute
hemothorax is usually identified by a fluid collection of homogeneous
hyperechogenicity, subsequently showing fluid-fluid levels as the blood
products separate.
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 13A —Chest radiographs of two infants. 5-month-old male. Image
illustrates displacement of mediastinal soft tissues and support apparatus to
right with intercostal drain decompressing left-sided hemothorax.
|
|
Sonography provides a useful tool for intracranial imaging in infants
during treatment and for documenting complications (Figs.
7A and
7B). Widening of the
interhemispheric fissure during ECMO is recognized in up to 59% of cases
[17,
20-22]
(Fig. 8). Typically, the
finding resolves once ECMO is discontinued
[20]. Some authors believe
this to be an intracranial manifestation of generalized edema
[20]. Other authors suggest
that raised sagittal sinus pressures associated with internal jugular vein
ligation may contribute to the appearance
[17,
23].
Complications of Treatment
Complications of ECMO are divided into mechanical and patient factors.
Given that ECMO is generally used as a rescue therapy in critically ill
infants and children, complication rates during treatment can be significant.
Typically, patients are severely hypoxemic, hypotensive, and acidotic before
commencement of ECMO and thus are already at particular risk of
cerebrovascular injury. Successful use of ECMO requires adequate
anticoagulation to prevent thrombus formation in the extracorporeal circuits,
but systemic heparinization places patients at significant risk of hemorrhagic
complications involving a range of organ systems. Furthermore, platelets are
continuously consumed during ECMO, with implications for coagulation. Platelet
transfusions are administered regularly to maintain an adequate platelet
count. Acute changes in hemodynamics during cannulation and in fluid balance
during therapy contribute to the risk of hemorrhage and fluid collections.
Several novel therapies have been used to minimize the risk of bleeding
complications. In North America, for instance, the use of an antifibrinolytic
drug, aminocaproic acid, in patients undergoing surgery before or while on
ECMO appears to significantly reduce the risk of surgical site bleeding
[24], although there is no
proven benefit on intracranial hemorrhage rates.
CNS Complications
Systemic anticoagulation makes intracranial hemorrhage the primary risk of
ECMO. Some authors have suggested CNS injury may be compounded by ligation of
the internal jugular vein and common carotid artery. Series suggest an
incidence of intracranial hemorrhage of approximately 14% during ECMO
[25,
26]. Both intracranial
hemorrhage and ischemic injury can be detected and monitored in infants by
cranial sonography. Particular attention should be paid to the posterior
fossa, where sonography views may be limited; and the echogenic tentorium,
vermis, and subarachnoid spaces may mask acute hemorrhage
[27]
(Fig. 9). Care must be taken
to examine the extraaxial spaces at the periphery of the field of view where
large extraaxial collections may be missed. This includes the routine use of a
high-resolution linear probe to examine the near field
(Fig. 10).
Although sonography remains a cornerstone in the monitoring of the CNS
during treatment, CT studies may provide additional information in up to 73%
of cases [17]. Generalized
edema, acute hypoxic ischemic injury, and small hemorrhages may not be visible
sonographically. In addition, cranial sonography may not be possible in older
patients because of closure of the fontanelle. In the neonatal population,
however, sonography remains highly sensitive for the detection of major
intracranial hemorrhage, an event that can affect the acute management of the
patient.
Thoracic Complications
Thoracic complications of ECMO are relatively common
[27] and include migration of
support apparatus, air collections, pleural effusions, and hemorrhage. A
proportion of complications may be unsuspected, and the clinical significance
of these sequelae varies widely. Chest sonography is often a useful adjunct to
portable radiography in such cases.
Migration of cannulae and chest drains can occur in small infants as
increasing soft-tissue edema displaces catheters that are sutured to the skin.
Intrathoracic air leaks can complicate management and are common due to the
underlying lung disease present in a large proportion of patients. Such
sequelae may be clinically apparent and are readily detected on chest
radiography (Figs. 11A and
11B). Large air collections
can dramatically compromise venous drainage and need to be addressed
urgently.
Hemorrhage into the pleural and mediastinal spaces is well recognized.
Sonography aids in the evaluation of pleural collections
(Fig. 12) and may show
hemorrhage, which appears echogenic and complex
[27,
28]. Although the lungs remain
opacified, significant mediastinal shift due to pleural collections or lung
collapse can be detected only by the displacement of the support lines and
tubing [29] (Figs.
13A and
13B). Occasionally, CT may
play a role in the detection of thoracic abnormalities such as pleural and
pericardial collections, mediastinal hemorrhage, lung abscess, and
bronchopleural fistulas [16].
It is important to stress, however, that complications detected during routine
examinations in stable patients may not require treatment until ECMO support
has been weaned; insertion of catheters or drains for air leaks or collections
may cause unnecessary and life-threatening hemorrhage.
Other Complications
Other less-well-recognized complications during ECMO include adrenal gland
hemorrhage [16,
30]; hepatic infarction
[16]; intraperitoneal
hemorrhage; and, occasionally, retinal hemorrhage
[31]. Rapid splenic
enlargement has been documented in a series of infants after commencement of
ECMO therapy [32]. The authors
of that study propose that splenomegaly may have been secondary to hemolysis
and platelet aggregation during ECMO support. In two cases, splenomegaly
impeded repair of a congenital diaphragmatic hernia. Periosteal reactions of
the ribs have been reported and are thought to be secondary to subperiosteal
edema during therapy [33].
Imaging Protocols
ECMO imaging protocols vary among centers. The importance of regular
surveillance for cannula migration and hemorrhagic complications cannot be
overstated but must be balanced by a regard for minimizing radiation dose. It
may be reassuring to remember that cannula position is routinely assessed
during regular echocardiography examinations performed while a patient is on
ECMO support. The current imaging protocol at our institution involves a
cranial sonography examination before commencement of ECMO as a baseline study
and to exclude a preexisting intracranial condition and again within 24 hr of
cannula insertion because of potentially significant changes in hemodynamics
and blood clotting after anticoagulation. Repeat cranial sonography
examinations are then performed at weekly intervals or in the event of
clinical deterioration. This protocol was established after an extensive
review of our practice that had shown a very low yield of significant new
intracranial findings after the first 24-hr period. However, different centers
have different protocols and there is no consensus in the literature; some
centers perform cranial sonography on a daily basis for the duration of the
admission.
Similarly, chest radiographs are obtained immediately before and after
commencement of ECMO and thereafter at 2- to 3-day intervals while the patient
remains stable because improvement in lung aeration is best assessed
clinically. The advent of anticipated pathways of recovery has led to a
further reduction in the number of examinations requested.
Conclusions
ECMO is a highly invasive treatment commonly used as a rescue therapy in
critically ill patients; complications are varied and frequent. The
radiologist must be able to recognize the normal radiographic appearances of
ECMO support and should be vigilant in the detection of complications. Given
the extensive range of cannulae currently available and the lack of
information provided by manufacturers regarding the radiographic appearances
of each catheter, close liaison with the clinical team is essential in sharing
information and allowing accurate assessment of catheter type and position.
Appropriate imaging protocols should be implemented to ensure maximal
efficiency and safety during therapy and, given the rapid expansion in this
field of life support, should be reassessed as practice evolves in individual
centers.
References
- [No authors listed]. UK collaborative randomised trial of neonatal
extracorporeal membrane oxygenation. UK Collaborative ECMO Trial Group.
Lancet 1996; 348:75
-82[CrossRef][Medline]
- Langham MR, Kays DW, Beierle EA, Chen MK, Stringfellow K, Talbert
JL. Expanded application of extracorporeal membrane oxygenation in a pediatric
surgery population. Ann Surg 2003;237
: 766-774[CrossRef][Medline]
- Kugelman A, Gangitano E, Pincros J, Tantivit P, Taschuk R, Durand
M. Venovenous versus venoarterial extracorporeal membrane oxygenation in
congenital diaphragmatic hernia. J Pediatr Surg2003; 38:1131
-1136[CrossRef][Medline]
- Dimmitt RA, Moss RL, Rhine WD, et al. Venoarterial versus
venovenous extracorporeal membrane oxygenation in congenital diaphragmatic
hernia: the Extracorporeal Life Support Organization Registry, 1990-1999.
J Pediatr Surg 2001;36
: 1199-1204[CrossRef][Medline]
- Fukuda S, Aoyama M, Yamada Y, et al. Comparison of venoarterial
versus venovenous access in the cerebral circulation of newborns undergoing
extracorporeal membrane oxygenation. Pediatr Surg Int1999; 15:78
-84[CrossRef][Medline]
- Streiper MJ, Sharma S, Dooley KJ, Cornish JD, Clark RH. Effects of
venovenous extracorporeal membrane oxygenation on cardiac performance as
determined by echocardiographic measurements. J
Pediatr 1993; 122:950
-955[Medline]
- Knight GR, Dudell GG, Evans ML, Grimm PS. A comparison of
venovenous and venoarterial extracorporeal membrane oxygenation in the
treatment of neonatal respiratory failure. Crit Care
Med 1996; 24:1678
-1683[CrossRef][Medline]
- Vogt JF, Gangitano EF, Liberman GR, et al. Cannulation of cephalad
segment of internal jugular vein during extracorporeal membrane oxygenation in
the neonate. Respir Care 2003;48
: 352-362[Medline]
- von Allmen D, Babcock D, Matsumoto J, et al. The predictive value
of head ultrasound in the ECMO candidate. J Pediatr
Surg 1992; 27:36
-39[CrossRef][Medline]
- De Sanctis, Bramson RT, Blickman JG. Can clinical parameters help
reliably predict the onset of acute intracranial hemorrhage in infants
receiving extracorporeal membrane oxygenation?
Radiology 1996;199
: 429-432[Abstract/Free Full Text]
- Reickert CA, Schreiner RJ, Bartlett RH, et al. Percutaneous access
for venovenous extracorporeal life support in neonates. J Pediatr
Surg 1998; 33:365
-369[CrossRef][Medline]
- Peek GJ, Firmin RK, Moore HM, Sosnowski AW. Cannulation of neonates
for venovenous extracorporeal life support. Ann Thorac
Surg 1996; 61:1851
-1852[Abstract/Free Full Text]
- Jamadar DA, Kazerooni EA, Cascade PN, et al. Extracorporeal
membrane oxygenation in adults: radiographic findings and correlation of lung
opacity with patient mortality. Radiology1996; 198:693
-698[Abstract/Free Full Text]
- Taylor GA, Lotze A, Kapur S, Short BL. Diffuse pulmonary
opacification in infants undergoing extracorporeal membrane oxygenation:
clinical and pathological correlation. Radiology1986; 161:347
-350[Abstract/Free Full Text]
- Schlesinger AE, Cornish JD, Null DM. Dense pulmonary opacification
in neonates treated with extracorporeal membrane oxygenation.
Pediatr Radiol 1986;16
: 448-451[CrossRef][Medline]
- Lidegran M, Palmer K, Jorulf H, Linden V. CT in the evaluation of
patients on ECMO due to acute respiratory failure. Pediatr
Radiol 2002; 32:567
-574[CrossRef][Medline]
- Bulas DI, Taylor GA, O'Donnell RM, Short BL, Fitz CR, Vezina G.
Intracranial abnormalities in infants treated with extracorporeal membrane
oxygenation: update on sonographic and CT findings. Am J
Neuroradiol 1996; 17:287
-294[Abstract]
- Tu CY, Hsu WH, Hsia TC, et al. Pleural effusions in febrile medical
ICU patients: chest ultrasound study. Chest2004; 126:1274
-1280[Abstract/Free Full Text]
- Kearney SE, Davies CW, Davies RJ, Gleeson FV. Computed tomography
and ultrasound in parapneumonic effusions and empyema. Clin
Radiol 2000; 55:542
-577[CrossRef][Medline]
- Rubin DA, Goss GW, Ehrlich SM, Alexander AA. Interhemispheric
fissure width in neonates on ECMO. Pediatr Radiol1990; 21:12
-15[CrossRef][Medline]
- Canady AI, Fessler RD, Klein MD. Ultrasound abnormalities in term
infants on extracorporeal membrane oxygenation. Pediatr
Neurosurg 1993; 19:202
-205[Medline]
- Babcock DS, Han BK, Weiss RG, Ryckman FC. Brain abnormalities in
infants on extracorporeal membrane oxygenation: sonographic and CT findings.
AJR 1989; 153:571
-576[Abstract/Free Full Text]
- Matamoros A, Anderson JC, McConnell J, Bolam DL. Neurosonographic
findings in infants treated by extracorporeal membrane oxygenation (ECMO).
J Child Neurol 1989;4
[suppl]: S52-S61
- Downard CD, Betit P, Chang RW, Garza JJ, Arnold JH, Wilson JM.
Impact of AMICAR on hemorrhagic complications of ECMO: a ten-year review.
J Pediatr Surg 2003;38
: 1212-1216[CrossRef][Medline]
- Khan AM, Shabarek FM, Zwischenberger JB, et al. Utility of daily
head ultrasonography for infants on extracorporeal membrane oxygenation.
J Pediatr Surg 1998;33
: 1229-1232[CrossRef][Medline]
- Hansell DR. Extracorporeal membrane oxygenation for perinatal and
pediatric patients. Respir Care 2003;48
: 352-366[Medline]
- Bulas DI, Taylor GA, Fitz CR, Revenis ME, Glass P, Ingram JD.
Posterior fossa intracranial hemorrhage in infants treated with extracorporeal
membrane oxygenation: sonographic findings. AJR1991; 156:571
-575[Abstract/Free Full Text]
- Gross GW, Cullen J, Kornhauser MS, Wolfson PJ. Thoracic
complications of extracorporeal membrane oxygenation: findings on chest
radiographs and sonograms. AJR 1992;158
: 353-358[Abstract/Free Full Text]
- Gross GW, Dougherty CH. Pleural hemorrhage in neonates on
extracorporeal membrane oxygenation and after repair of congenital
diaphragmatic hernia: imaging findings. AJR1995; 164:951
-955[Abstract/Free Full Text]
- Sivit CM, Short BL, Revenis ME, Rebolo LC, Brown-Jones C, Garin DB.
Adrenal hemorrhage in infants undergoing ECMO: prevalence and clinical
significance. Pediatr Radiol 1993;23
: 519-521[CrossRef][Medline]
- Pollack JS, Tychsen L. Prevalence of retinal haemorrhage in infants
after extracorporeal membrane oxygenation. Am J
Ophthalmol 1996; 121:297
-303[Medline]
- Klippenstein DL, Zerin JM, Hirschl RB, Donn SM. Splenic enlargement
in neonates during ECMO. Radiology 1994;190
: 411-412[Abstract/Free Full Text]
- Feinstein KA, Fernbach SK. Periosteal reaction of the ribs in
neonates treated with extracorporeal membrane oxygenation: prevalence and
association with soft-tissue swelling. AJR1993; 160:587
-589[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
