Although sedation is necessary for most MR studies of children younger than 8 years, fast MR techniques such as EPI, half-Fourier acquisition single-shot turbo spin echo (HASTE, also called single-shot fast spin echo or ssFSE by manufacturers) and rapid acquisition with relaxation enhancement (RARE, also called fast spin echo, FSE, or turbo spin echo, TSE) (5,6) allow high-quality images of the entire brain to be obtained within a sufficiently short examination time (<1 minute) to avoid significant motion artifacts. Alternatively, the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER or PROP) acquisition and reconstruction scheme is relatively tolerant of patient motion (7,8). EPI and HASTE/ssFSE are often adequate for
diagnosis of ventriculomegaly/shunt failure or acute infarcts but are not, at this time, of adequate quality to allow diagnosis of subtle malformations, leukodystrophies, or tumor progression. PROPELLER will allow identification of malformations and tumors but does not have the contrast resolution to detect subtle white matter abnormalities.

If sedation can be avoided, it should be, as complication rates are generally in the range of 0.4% to 1% even in the best of hands (9). The incidence of complications seems to be significantly higher if the child has a history of respiratory illness or if more than one agent is used for sedation (9). Therefore, it is important to remember that sedation is not universally necessary for obtaining standard MR sequences of pediatric patients. Neonates can often be scanned without sedation if they are fed immediately prior to the study, kept warm, and earmuffs/ earplugs are used to reduce the noise; we have found the use of 24% oral sucrose (Sweet Ease) and nonnutritive sucking (i.e., a sucrose pacifier) immediately prior to the scan to be useful by promoting a calm state and reducing stress. Mathur et al. report considerable success in scanning unsedated neonates (10). We have had success with this, as well, as described in the section below on Special Issues in Scanning of Premature Neonates. Scanning neonates requires an armamentarium of MR compatible life-support and monitoring equipment, trained transport personnel, at least 1 hour of preparation time before the baby is transported to the scanner, and a similar amount of time after the scan (10).

With adequate personnel and equipment, it can be achieved with a high success rate. The success rate diminishes with the maturity of the baby, however. As they mature, infants become increasingly aware of their surroundings and sedation becomes increasingly necessary in order to obtain diagnostic scans; by the age of 2 to 3 months, sedation is nearly always necessary. Fewer older children need sedation if some sort of MR-compatible audiovisual system is available. We find that most children aged 8 years and older as well as some 6- and 7-year-olds can hold still for an MR study if they are adequately enraptured by a movie. Harned and Strain (11) found a 25% reduction in the number of sedations for children aged 3 to 10 years and a 50% reduction in sedations of children more than 10 years old when they began to use an audiovisual system in their MR scanner. They also found a 17% reduction in the MR room time for patients who were examined without sedation. Hallowell et al. (12) prepared children (aged 3.7-17 years) in a practice MRI unit (without real magnets); of the 78% who performed adequately during preparation and underwent an MR scan, they were able to obtain diagnostic noncontrast images in 96%. However, contrast-enhanced images were not obtained and motion artifact was seen in more than 60% of studies.

When sedation is necessary, the practitioner should be familiar with the guidelines published by the American Academy of Pediatrics (13,14), the American Society of Anesthesiologists (15), and the American College of Radiology (16). These guidelines require the following provisions:

In addition, the 2008 update of the AAP guidelines requires the following (14); please read all of these guidelines before embarking on pediatric MRI sedation.

Although the most common complications from sedation are related to respiratory compromise, special attention must be given to monitoring body temperature in the neonatal and the young pediatric population. Temperature monitoring equipment that is approved for use in the MR suite is becoming more readily available. A growing number of commercial MR-approved neonatal isolation transport units and other warming devices have also become more readily available for use during MR scans.

A number of different drugs can be and have been used safely for pediatric sedation (17,18,19,20,21,22,23,24,25,26,27). Specific anesthetic regimens vary regionally and institutionally, and a comprehensive discussion of appropriate medications for sedation during imaging is beyond the scope of this text. Some regimens that have been used safely by physicians credentialed in pediatric sedation in our practice are listed below for reference. The reader should consult with local anesthesiologists and/or pediatricians before employing any anesthesia protocol, whether listed here or not. Moreover, close contact should be maintained with an anesthesiologist or experienced pediatrician for consultation in difficult cases.

Sedation can be administered by many routes: oral (PO), rectal (PR), intramuscular (IM), or intravenous (IV). PO, PR, and IM routes have the important advantage that they do not require special skills for administration. However, the absorption of many drugs from the
muscle and gastrointestinal tract can be erratic. Thus, sedatives and anesthetics administered in large doses by these routes can be difficult to reverse in the event a complication arises during sedation. The recommendation of the American Academy of Pediatrics is that infants should be given nothing by mouth for at least 4 hours prior to deep sedation; older children for at least 6 hours (28). This accelerates bioabsorption in the case of PO and PR medications but, more importantly, decreases the risk of pulmonary aspiration during the examination.

Sodium pentobarbital (Nembutal) is a useful short-acting barbiturate for enteral or parenteral sedation in pediatric patients (20), although it should be used with care in patients who have hepatic or metabolic disease, as the drug may alter the metabolism of other medications. The usual dose of sodium pentobarbital is 6 mg/kg for the first 15 kg of body weight, followed by 5 mg for each additional kg up to a total dose of no more than 100 mg (as a single IM injection). The drug should be administered 35 to 45 minutes before the anticipated imaging time.

Historically and in a previous edition of this book, the safest drug for sedating children was considered to be chloral hydrate. Others have also reported on the safety and efficacy of chloral hydrate, especially in infants under the age of 18 months (22) and even up to age 4 years (29). However, because questions have been raised concerning the potential carcinogenicity of this drug (30), its use has been restricted in some countries. In the United States, the American Academy of Pediatrics has judged the evidence insufficient (31), and many physicians continue to use chloral hydrate without incident. Usually administered orally, known side effects include prolonged somnolence (>4 hours), unsteadiness, and hyperactivity in some children after waking (32). Such symptoms may require the child to be kept at the imaging center and observed for a prolonged time period. In addition, one study suggested that the complication rate, particularly for episodes of oxygen desaturation, is lower with oral pentobarbital than with chloral hydrate and recommended that oral pentobarbital be used as an equally effective and safer drug than chloral hydrate in infants (33). If used, chloral hydrate is administered orally in a dose of 50 to 75 mg/kg for the first 10 kg of body weight, then 50 mg/kg for each kg of body weight above 10 kg. If the child is still awake after 20 minutes, supplementary doses can be given up to a total dose of 2000 mg. Chloral hydrate is not currently used for infant sedation at UCSF.

After enteric or intramuscular sedatives have been administered, the patient and parent should be transferred to a dark, quiet room. The noise and activity of a busy waiting room will otherwise keep the child awake. Sleep deprivation is also helpful. If the imaging study is performed in the morning, keep the child awake several hours past normal bedtime the night preceding the exam and awaken the child earlier than usual on the morning of the exam. If the imaging is performed in the afternoon, deprive the child of his or her normal naps. Do not allow the child to fall asleep during travel to the imaging facility. Finally, permit a parent to enter the scanner with the patient, if desired, in order to reassure the child that security is nearby. The second body will not create artifact, as long as the parent has been screened adequately for prosthetics or other metallic objects.

IV sedation offers a number of advantages over other routes of administration. Once IV access is obtained, sedation can be rapidly achieved. Dosage can be titrated more easily, and supplemental doses are easily administered without disturbing the patient. Finally, and perhaps most importantly, the effects of many drugs (e.g., opiates and benzodiazepines) can be rapidly reversed. Reversal agents should be readily available for all cases in which sedation is administered. Flumazenil (Romazicon), a benzodiazepine receptor agonist, is administered in IV doses of 0.01 mg/kg (up to 0.2 mg) over 15 seconds and repeated up to a maximum of four additional times. Naloxone (Narcan), in contrast, is used for reversal of respiratory depression from opioids and given intravenously in increments of 0.005 to 0.01 mg at 2 to 3 minute intervals. It is important to keep in mind that the half-life of both reversal agents is less than most of the drugs that they are used to reverse, and that patients should thus be monitored following the procedure in the event that additional doses of these agents are necessary.

The most popular drug used for IV sedation is pentobarbital (21,22), which is administered intravenously in the following manner. Approximately 2.5 mg/kg is given over 30 to 40 seconds while closely observing the patient. If the patient does not fall asleep within 60 seconds, a second dose of 1.0 mg/kg is administered. If the patient remains awake, a third and fourth administration of 1.0 mg/kg can be given up to a total dose of 6 mg/kg. Repeated doses of 1 to 1.25 mg/kg are then administered, as needed, to maintain sedation during the exam. Small motions of the arms and legs are usually an indication that the level of sedation is diminishing. At UCSF, our pediatric anesthesiologists have recommended that pentobarbital be used in conjunction with morphine sulfate, as the two drugs have complimentary actions. This regimen consists of an initial dose of 1 to 2 mg/kg pentobarbital, followed by 0.05 mg/kg morphine, followed by another dose of pentobarbital, another of morphine, etc., until the patient is adequately sedated. Additional doses are given, as needed, if the level of sedation appears to be diminishing.

IV sedation using propofol (Diprivan), an ultrashort-acting sedative-hypnotic, has been successfully used instead of pentobarbital in some studies (26). Propofol acts rapidly and, more importantly, is rapidly metabolized so that the sedated children rapidly recover after imaging. The disadvantage of this drug is the significantly higher rate of respiratory events compared to, for example, pentobarbital, requiring significantly more airway manipulations to relieve obstruction (34). Therefore, in the United States, it has been recommended that only anesthesiologists administer propofol (15). As a result, those wishing to take advantage of the rapid induction and recovery of pediatric patients sedated with propofol need to arrange to have pediatric sedation performed by anesthesiologists or nurse anesthetists. This is currently the situation at UCSF. The reader should be aware that the use of propofol with 100% oxygen can cause artifacts in fluid attenuation inversion recovery (FLAIR) images, with the cerebrospinal fluid (CSF) appearing abnormally hyperintense in the sulci and cisterns (35,36). If CSF hyperintensity is seen in the cisterns on FLAIR sequences in a sedated child, the agent used for sedation and the percent of supplemental oxygen must be researched before the scan is interpreted as abnormal. Using lower concentrations of supplemental oxygen (50%-60% works well) eliminates this artifact (37,38).

CT scans in children should initially be performed without IV contrast. If a noncontrast scan reveals an abnormality and an MR cannot be obtained in a timely fashion, IV contrast should be given. If the noncontrast CT scan is normal, very little information is gained by administering contrast unless a vascular lesion is suspected (49). Nonionic, iso-osmolar (Visipaque) or low-osmolar (Omnipaque) contrast media should be used because they are safer and less uncomfortable for the patient. The exact type of iodinated contrast is not important, as long as the concentration of iodine is approximately 300 mg/mL. The recommended dose is 3 mL/kg of body weight up to a total dose of 120 mL. The child should be scanned as soon as possible after the contrast has been administered. Adverse reactions to iodinated contrasts are rare in the pediatric population. They are most rare in the youngest patients (50). Acute reactions are most common in children weighing 24 to 40 kg. Asthma and previous reactions to contrast medium are risk factors for acute reactions (50).

MRI can be performed in most cases without paramagnetic MR contrast agents, although the administration of gadolinium chelates may improve the identification and evaluation of primary and metastatic brain tumors (especially extraparenchymal tumors), infections (abscess, empyema, cerebritis, and meningitis), neoplasms of the spinal cord and spinal canal, and some neurocutaneous disorders (neurocutaneous melanosis, neurofibromatosis type II, Sturge-Weber syndrome) (51,52) and permits the assessment of cerebral perfusion using dynamic susceptibility contrast MRI (DSC-MRI) methods (53,54). Administration of contrast to infants or children with developmental delay is unlikely to be of any diagnostic benefit unless a space-occupying lesion is identified on a noncontrast study or a cutaneous lesion suggesting one of the neurocutaneous disorders mentioned above is present. Similarly, no advantage is gained by administering paramagnetic contrast agents to children with malformations of the brain or spine, with the exception of dermal sinus tracts (see Chapter 9).

Once widely administered in the setting of impaired renal function, gadolinium chelates are now known to be associated with nephrogenic systemic fibrosis (NSF) in patients with renal insufficiency (55,56). This rare but serious systemic disorder causes fibrosis of the skin and other tissues, thereby leading to considerable morbidity or even death. To the date this chapter was revised; there have been nine reported cases of NSF in children after gadolinium injection (57,58). All of these occurred in patients with low glomerular filtration rate (GFR < 30), although there are a few case reports of NSF in adults with slightly higher GFR between 40 and 60. A proinflammatory state (e.g., systemic infection, limb or major tissue injury, recent surgery or thrombosis) may increase the risk of developing this complication. With the recognition of the relationship between gadolinium and NSF, there is no longer a justification for administering contrast to every sedated patient “just in case,” that is, in order to avoid resedating the
child for a repeated contrast-enhanced scan. When gadolinium is necessary, screening for NSF risk factors is essential, and the relative risks and benefits in patients with renal failure should be carefully considered in consultation with referring physicians.

At high and repeated doses, IV gadolinium is teratogenic in animal studies (59). Although no similar effects have been observed in human studies of teratogenicity, gadolinium is known to cross the placenta and may be excreted in the amniotic fluid, where the dwell time of the agent may exceed several days. There are no indications for which maternal paramagnetic contrast agent administration is necessary for fetal MRI.

From a diagnostic efficacy standpoint, no significant difference has been demonstrated among the different paramagnetic contrasts that are commercially available. All are given intravenously; the standard dose is 0.1 mmol/kg (corresponding to a volume of 0.2 mL/kg for nearly all commercial preparations). After infusion of contrast, only T1-weighted spin-echo, three-dimensional Fourier transform (3DFT) gradient echo, (i.e., spoiled gradient recalled [SPGR], or magnetization prepared rapid acquisition gradient echo [MPRAGE]), FLAIR, or dynamic contrast-enhanced MRI perfusion sequences should be obtained; fat suppression may be useful if meningeal disease is suspected (60) or if suspected pathology lies in the orbits or neck, where fat may obscure the enhancing lesion. T2-weighted images have no value in this situation. It is likely that higher doses of contrast (0.3 mmol/ kg) or the use of magnetization transfer pulses to suppress white matter signal allows a higher sensitivity for detection of enhancing lesions, particularly CSF-borne metastases from brain tumors (4); however, higher doses likely increase the incidence of NSF and, therefore, should be used only when the benefits of added contrast, such as identification of a condition that might alter therapy, outweigh the potential risks.

A sagittal, T1-weighted sequence should be performed on all patients. This sequence allows critical assessment of the midline structures that are frequently abnormal in congenital brain malformations. Sagittal images are optimal for evaluating the corpus callosum, pituitary gland, hypothalamus, and cerebellar vermis, common locations of pediatric brain tumors. They are also excellent for assessing the lateral convexities of the cerebral hemispheres, especially around the sylvian fissures. When acquired as the first sequences in the examination, this sequence can also be used as a localizer for additional axial or coronal sequences. Using a standard quadrature head coil at 1.5 T, suitable parameters for T1-weighted imaging are a repetition time (TR) of 500 to 600 milliseconds, echo time (TE) as short as possible (20 milliseconds or less), 3 to 4 mm slice thickness (1 mm gap), acquisition matrix 192 × 256, and one signal average. Using these parameters, it is possible to cover most of the brain in the sagittal plane. If the option of 192 acquisitions in the phase encoding direction is not available, 128 phase encodings (at the expense of decreased spatial resolution) or 256 phase encodings (at the expense of increased imaging time and diminished signal-to-noise ratio) can be used. Using a short TR (<500 milliseconds) will save imaging time while still allowing assessment of the midline structures, but a price is paid in that signal-to-noise decreases and multislicing may not allow the lateral portions of the brain to be covered unless fractional k-space acquisitions or a rectangular field of view is also employed. If a phased-array head coil and parallel imaging are available, particularly at 3 T, a 3D radiofrequency or gradient-spoiled gradient-echo volumetric sequence (SPGR, MPRAGE, 3D-FLASH, CE-FFE-T1, T1-FAST) should be obtained in the sagittal plane (employing the largest, craniocaudal dimension as the readout axis for quickest acquisition) and high-quality reformations in the coronal and axial (and, if necessary, true sagittal) planes can be easily obtained. This has become our most useful sequence for the diagnosis of structural abnormalities. For this sequence, we use TR/TE = 35 milliseconds/minimum, flip angle 35°, 22 × 22 cm FOV, 1 mm partition size, and 256 × 192 acquisition matrix with no signal averaging. If imaging at 3 T is not available or thin volumetric images are not desired, FLAIR T1 acquisition using TR/TE = 600/20 milliseconds and inversion time (TI) of 750 milliseconds also gives excellent T1-weighted images.

After the sagittal T1 images have been obtained, axial T2-weighted images should be acquired in all patients. For patients older than 24 months, only these T2-weighted images are necessary for the most basic imaging protocol. If a volumetric sequence has not been acquired and reformatted axially in infants with developmental delay or seizures, we also obtain an axial T1-weighted sequence (inversion recovery, spin echo). For patients less than 18 months of age, both T1- and T2-weighted images are necessary for accurate interpretation. Brain maturation is evaluated best by T1-weighted images from birth to 6 months of age. However, from 6 to 8 months of age until approximately 24 months of age (at which time the brain is essentially mature by MR standards), T2-weighted images are more useful for assessing myelination and brain maturity. During the process of white matter maturation, the cerebral cortex and subcortical white matter of the brain become isointense for a variable period of time on MR images; this isointensity obscures structural detail (see Chapter 2). Therefore, during the first 8 months of life (while the white matter is maturing on T1-weighted images), T2-weighted images are necessary to see the details of the gyral and sulcal patterns. Similarly, as the white matter matures on T2-weighted images between 8 and 24 months of age, T1-weighted images are essential for evaluation of structural abnormalities.

The optimal imaging parameters may vary from one scanner to another and at different field strengths. 3DFT spoiled gradient-echo techniques that use radiofrequency or gradient pulses to “spoil” residual transverse magnetization have the ability to acquire very thin (1 mm or less) contiguous images through the brain in a relatively short imaging time. These techniques are particularly useful in looking for small, subtle cortical malformations (see Chapter 5) and for subtle areas of T1 shortening (injured brain) in neonates and young infants (see Chapter 4). As discussed earlier, they also have the important advantage that they
can be reformatted in any plane (sagittal, axial, coronal, or any obliquity), allowing acquisition of only one T1-weighted sequence. We routinely obtain volumetric gradient-echo sequences in all neonates, all patients with developmental delay, and patients being imaged for epilepsy. For brain tumors, neuronavigation techniques are used at our institution, and these require volumetric techniques as well, for T1-weighted images. As visualization of enhancement is important in these patients, we use slightly different parameters for postgadolinium imaging, with TR/TE = 34 milliseconds/minimum, flip angle 70°, and partition size of 3 mm. If volumetric images cannot or are not acquired, T1 FLAIR or spin-echo sequences can be used. Although T1 FLAIR gives better T1 information than conventional spin-echo sequences, an SE 600/minimum sequence (similar to that outlined for the sagittal images) can be routinely used for T1-weighted images on 1.5 T scanners. Spin echo is the sequence of choice when paramagnetic contrast is to be administered. Use of this imaging sequence produces excellent images; moreover, the imaging time is shorter than with most inversion recovery sequences, resulting in less motion artifact. Spinecho T1 images do not work as well at 3 T due to the prolongation of T1 relaxation times, however. Therefore, T1 FLAIR (usually using short TEs in the range of 20 milliseconds and TIs of 750 milliseconds) sequences are used; these give excellent T1 weighting but do not show contrast enhancement as well as spin-echo images or spoiled gradientecho images.

For T2-weighted sequences, dual spin-echo sequences with a 4 to 5 mm slice thickness (2-2.5 mm gap), TR of 2500 to 3000 milliseconds, TE of 30 to 60 milliseconds (first echo) and 70 to 120 milliseconds (second echo), 192 × 256 acquisition matrix, and 0.75 to 1 excitation are recommended. In infants less than 12 months old, heavily T2-weighted sequences are highly recommended, as the water content of the brain in young children is considerably higher than in older children and adults (76,77). We routinely use TR = 3000 milliseconds and TE = 60, 120 milliseconds (first and second echoes) for this patient group. Others use a dual-echo short tau inversion recovery (STIR) sequence with TR = 5400 milliseconds, TI = 130 milliseconds, and TE = 30, 128 milliseconds (52). Although RARE (also called fast spin echo [FSE] or turbo spin echo [TSE]) images are used by some groups for infants and children and seem to show myelination with fair accuracy (78), we and others (52) are not completely satisfied with the contrast between gray and white matter of commercially available RARE sequences, especially with long echo train lengths. Contrast between gray and white matter, and between normal and pathologic tissues, is less in infants than in older children and adults; therefore, imaging parameters and sequences must be adjusted to maximize contrast. Moreover, although missing a single small white matter lesion in an adult is relatively unimportant, the same type of abnormality may be the only clue to the correct diagnosis in a child (79). Therefore, we continue to use conventional spin-echo technique in infants with nonspecific developmental delay. However, we do use RARE sequences to supplement our standard images; they are ideal for obtaining T2-weighted images in a second plane for infants with focal neurologic signs or symptoms and for evaluation of congenital anomalies or tumors (80). In addition, children who are imaged without sedation can benefit from the shorter acquisition time of RARE sequences. Typical parameters include TR = 2500 to 3000 milliseconds (up to 4000 milliseconds in infants <12 months old), effective TE = 17 to 100 milliseconds, and echo train length = 8. We also use 3D RARE sequences to assess for subtle cortical anomalies in infants at times when subcortical white matter is isointense to cortex on our volumetric T1-weighted images (usually between ages 3 and 8 months). For this technique, we use TR = 4000 milliseconds, TE = 85 milliseconds, echo train length = 16, partition size = 1.5 mm, and a 192 × 256 acquisition matrix.

Low-field scanners (≤1.0 T) do not produce adequately T1-weighted images with conventional spin-echo technique. The reasons for the poorer images are twofold. First, T1 relaxation times increase with increasing static magnetic field strength; therefore, images obtained with identical parameters will be more T1-weighted at higher field strengths than at lower field strengths. Second, lower field strength scanners, in general, are unable to achieve TEs as short as high-field scanners. The combination results in unsatisfactory T1-weighted spinecho images in infants scanned on most midfield and low-field scanners. Inversion recovery or 3DFT GE images should be used instead. The parameters for these sequences are similar to those used at 1.5 T, although the flip angle and TR in particular may require fine-tuning in order to maximize gray-white contrast.

Some authors have advocated the use of T2 FLAIR sequences to look for regions of abnormal T2 prolongation. In our experience and that of others (81), FLAIR images are not sensitive to cerebral pathology in neonates or infants but are useful in older children in whom myelination is complete or nearly complete. We do not use FLAIR as a primary sequence in infants, but we do sometimes use it as a secondary sequence for lesion characterization. In children beyond the age of 2 years, when myelination is nearly complete, FLAIR becomes a part of our routine protocol because of its high sensitivity for subtle supratentorial lesions, particularly in the cerebral cortex and periventricular white matter. Sargent and Poskitt found that FLAIR is complementary to T2-weighted infants and children (82). They found fast spin-echo FLAIR to have better CSF nulling and better gray matter-white matter differentiation than echo-planar FLAIR. Our standard parameters for 2D FLAIR imaging are TR/TE = 10,000/140 milliseconds, TI = 2200 milliseconds, with 3 mm contiguous slices and 192 phase encoding steps. More recently, we have started to use single slab 3D FSE FLAIR sequences (CUBE, SPACE, VISTA, depending on the manufacturer), which combine parallel imaging and variable flip angle excitation to convert slowly decaying longitudinal magnetization back to transverse magnetization. These have the benefit of very high spatial resolution of 1 to 2 mm and relatively short imaging times of 7 to 8 minutes, and like volumetric gradient-echo T1 images can be reformatted in any plane. However, longer imaging times render these images more susceptible to artifacts from patient motion, and in our experience the tissue contrast of these sequences does not always compare favorably to 2D FLAIR sequences.