Đề ôn thi thử môn hóa (703)

e5

164. Horvat CM, Ogoe H, Kantawala S, et al. Development and performance of electronic pediatric risk of mortality and pediatric logistic
organ dysfunction-2 automated acuity scores. Pediatr Crit Care
Med. 2019;20(8):e372-e379.
165. Gupta P, Rettiganti M, Gossett JM, Daufeldt J, Rice TB, Wetzel
RC. Development and validation of an empiric tool to predict favorable neurologic outcomes among PICU patients. Crit Care Med.
2018;46(1):108-115.
166. Watson DS, Krutzinna J, Bruce IN, et al. Clinical applications of machine learning algorithms: beyond the black box. BMJ. 2019;364:l886.
167. Wiens J, Saria S, Sendak M, et al. Do no harm: a roadmap for responsible machine learning for health care. Nat Med. 2019;25(9):
1337-1340.
168. Ginestra JC, Giannini HM, Schweickert WD, et al. Clinician perception of a machine learning-based early warning system designed
to predict severe sepsis and septic shock. Crit Care Med. 2019;
47(11):1477-1484.

169. Eytan D, Jegatheeswaran A, Mazwi ML, et al. Temporal variability
in the sampling of vital sign data limits the accuracy of patient state
estimation. Pediatr Crit Care Med. 2019;20(7):e333-e341.
170. Fartoumi S, Emeriaud G, Roumeliotis N, Brossier D, Sawan M.
Computerized decision support system for traumatic brain injury
management. J Pediatr Intensive Care. 2016;5(3):101-107.
171. Jha RM, Elmer J, Zusman BE, et al. Intracranial pressure trajectories: a novel approach to informing severe traumatic brain injury
phenotypes. Crit Care Med. 2018;46(11):1792-1802.
172. Sorani M, Hemphill J, Morabito D, Rosenthal G, Manley G. New
approaches to physiological informatics in neurocritical care. Neurocrit Care. 2007;6:1-8.
173. Hemphill J, Barton C, Morabito D, Manley G. Influence of data
resolution and interpolation method on assessment of secondary
brain insults in neurocritical care. Physiol Meas. 2005;26:373-386.


e6

Abstract: Much of the practice of brain-directed critical care in
children is empiric, but studies in traumatic brain injury show
that protocol-directed care with multidisciplinary teams in the
intensive care unit improve outcome. Neuromonitoring incorporates
the use of technologies including electroencephalography and
neuroimaging but begins with a focused neurologic examination.
Communication between team members, serial examinations,
anticipation and early recognition of changes in the neurologic

exam, or other monitoring parameters are essential. The examiner
should focus both on localization of a neurologic deficit and identifying mechanism(s) of injury (or potential injury) as the first
step in developing a treatment approach to reduce brain injury.
Key words: electroencephalogram, transcranial Doppler, intracranial pressure, neurologic examination, near infrared spectroscopy,
quantitative EEG


61
Neuroimaging
FRANCISCO A. PEREZ, HEDIEH KHALATBARI, AND DENNIS W.W. SHAW

PEARLS
• Multiple imaging modalities are available for evaluation of the
brain, head, neck, and spine of the critically ill child. The most
appropriate modality depends on consideration of patient pretest probability for the clinically suspected diagnosis, the modality sensitivity, and the patient’s age and condition.
• When ordering radiographic studies, particularly computed
tomography scans, the physician should keep the radiation

burden in mind, especially for infants. When ordering magnetic
resonance imaging for young children, one must keep in mind

the risk related to sedation and general anesthesia.
• The ever-increasing complexity of imaging modalities and medical problems in the intensive care unit warrants liberal consultation with radiology colleagues to yield an appropriately tailored imaging protocol and a more relevant interpretation.

Imaging Modality Overview

moderate to severe white-matter injury. It is less sensitive for pathologies such as mild to moderate parenchymal ischemic changes,
subarachnoid and punctate parenchymal hemorrhage, dural venous sinus thrombosis, and cerebral malformations. Extraaxial
collections, especially those along the lateral cerebral convexities,
can be difficult to visualize on ultrasound performed through a
midline fontanelle.
Color Doppler ultrasound uses the shift in frequency associated with reflection of the sound beam off a moving interface (the
“Doppler shift” phenomenon) to detect motion in the image
field, most commonly from blood in vessels. Those pixels with
movement are assigned a color to distinguish them from pixels
without movement. The color assigned (most often red and blue)
depends on whether movement is away or toward the transducer
and does not necessarily indicate arterial or venous flow. Color
Doppler allows for some investigation of the cerebral circulation,
primarily through the open fontanelle. Transcranial Doppler
(TCD) uses the same Doppler shift to produce waveforms that
give information about flow velocity and direction. An advantage
of TCD is that it can be performed through the thinner portions
of the skull, primarily the temporal squamosa, including in older
children and adults (Fig. 61.2). TCD has been used in the evaluation of cerebral perfusion in patients with sickle cell disease1 and
vasospasm secondary to subarachnoid hemorrhage (SAH).2 TCD
is generally sensitive to vessel stenosis greater than 50% in the
central cerebral circulation, with the highest sensitivity and specificity being in the middle cerebral artery (MCA).3,4 The role of
TCD in the ICU is expanding because it enables noninvasive
bedside monitoring of cerebral perfusion parameters in various
conditions,5 including in patients on extracorporeal membrane

oxygen therapy.6 However, the utility of TCD in critically ill children requires additional investigation. Moreover, performance of
TCD requires specially trained technologists and interpreting
physicians.

Multiple imaging modalities are available to investigate the neurologic status of children in the intensive care unit (ICU). To select the most appropriate imaging modality, many factors need to
be considered, including the pathologies suspected in a particular
case, sensitivity of the imaging modality for the suspected pathologies, complexity of the imaging examination, and the child’s
clinical condition. This chapter reviews available imaging modalities and the disease processes that are more commonly evaluated
with neuroimaging.

Radiography
The most commonly performed radiographic study in neuroimaging is the radiographic shunt series. The cerebrospinal
fluid (CSF) shunt is imaged in the frontal and lateral projections to evaluate the location and course of the proximal catheter, reservoir, valve, and distal catheter in order to detect
complications such as disconnection, fracture, improper placement, and migration (Fig. 61.1).

Ultrasound
Standard cranial ultrasound is a mainstay in the neonatal ICU.
Ultrasound generates grayscale digital images using sound waves
reflecting off tissue interfaces. There is no ionizing radiation and
bedside imaging can be performed. Because sound waves cannot
readily penetrate the bones of the skull, cranial ultrasound has
a restricted window through which sound waves can be used to
visualize the brain, primarily through an open anterior fontanelle.
Therefore, it is mostly limited to the first few months of life.
Ultrasound is sensitive to the detection of germinal matrix
hemorrhage, assessment of ventricular size, and identification of

735

736

S E C T I O N V I   Pediatric Critical Care: Neurologic

A

B

• Fig. 61.1  ​Improper placement of distal catheter of a cerebrospinal fluid (CSF) shunt in the large bowel.

Fourteen-year-old female with history of multiple prior CSF shunt revisions and intraperitoneal adhesions
underwent CSF shunt revision. Anteroposterior abdominal radiograph (A) from a radiographic shunt series
obtained the same day demonstrates the distal catheter of a ventriculoperitoneal shunt coursing along the
descending colon and sigmoid colon (arrows) and pneumoperitoneum (best appreciated in the right upper
and lower quadrants, asterisk). Coronal image (B) from a subsequent abdominal computed tomography
scan confirms the improper placement of the distal catheter in large bowel (arrow).

Computed Tomography
Computed tomography (CT) scanning has higher sensitivity
compared with ultrasound for many intracranial pathologies, including most neurosurgical emergencies, and is not limited by
closing and closed fontanelles. Modern scanners with multiarray
detectors image the patient rapidly (decreasing the need for sedation) and acquire volumetric data that may be reformatted in any
plane. CT, however, requires transporting the patient from the
ICU (although some portable units are available) and uses ionizing radiation (x-rays) to produce digital computer-reconstructed
images based on differences in tissue density (which affects x-ray
attenuation). The choice of radiation parameters, slice thickness,
postprocessing, and image viewing (window width and level)
must be tailored to the particular clinical question to optimize the
images. For example, CT may be performed at a lower radiation
dose when the clinical indication is to evaluate ventricular size.

However, this lower-dose CT may not be sensitive to parenchymal abnormalities. Therefore, the ICU team and radiologist need
to closely communicate in order to tailor the study appropriately.
With advancements in available computational power, newer CT
scanners employ complex postprocessing algorithms, such as iterative reconstruction, which enable scanning with lower radiation doses while maintaining image quality. New CT neuroimaging techniques—such as dual-energy CT, in which different
energy x-ray sources are simultaneously used—allow improved
differentiation of iodine, calcium, and hemorrhage; automated
bone removal in CT angiography and CT venography; production of virtual nonenhanced CT images; and metal artifact reduction in the instrumented spine.7,8
CT uses differences in density to generate images. Bone and
other calcifications have the highest density. Other components in

decreasing density are soft tissue (e.g., brain), water (e.g., CSF),
fat, and air. On CT, increasing density is represented by an increase in brightness on the images. However, the human visual
system is not able to appreciate the entire range of image intensities that are necessary to represent the full range of density values
measured by CT. Therefore, the CT images have to be reviewed
while adjusting window and level settings to interrogate ranges of
density. The difference in density between bone and brain is great,
whereas the difference between brain and fat is smaller. The density difference between gray and white matter is much smaller but
is sufficient to be appreciated with the appropriate window and
level. Generally, as edema develops in nonfatty tissue, there is a
decrease in density. Acute blood (approximately between 12 and
72 hours old) is of higher density than brain, and CT is sensitive
in the detection of acute (generally less than 7–10 days old) parenchymal, intraventricular, and extraaxial hemorrhage (epidural,
subdural, and subarachnoid; Fig. 61.3). CT is also useful in the
evaluation of ventricular size, extraaxial collections, cerebral
edema, mass effect, and cerebral herniation (secondary to spaceoccupying mass or diffuse cerebral edema).
Iodinated intravenous (IV) contrast can be used with CT to
evaluate the integrity of the blood-brain barrier (BBB) and differences in tissue perfusion. With injection of IV contrast, vessels
demonstrate a transient increase in density, with a variable rate of
contrast leak into the tissues depending on the local pathophysiology. The rate of equilibration of contrast concentration between
vessels and parenchyma is much slower in the brain because of the

intact BBB. Contrast then is especially useful in the brain, where
disruption of the normal BBB can reveal underlying pathology.
This increased density from contrast material is also used to image
arteries and venous structures with CT angiography (CTA) and
CT venography (Fig. 61.3C). With the appropriate software,


CHAPTER 61  Neuroimaging

737

B

A

C

D

F

•  Fig. 61.2  ​Transcranial Doppler ultrasound images. Color and spectral

E

Doppler waveforms from (A) normal right middle cerebral artery (MCA)
and (B) right terminal internal carotid artery (ICA). Abnormal right middle
cerebral (C) and terminal internal carotid (D) waveforms in another patient
with sickle cell disease and prior history of stroke. Magnetic resonance
angiographic images (E and F) of this patient with abnormal transcranial

Doppler demonstrate narrowed proximal MCAs and terminal ICA.