|
|
Features Departments Information |
  Sharon E. Byrd, MD
 Mary Ann Radkowski, MD
Crystal F. Darling, MD
|
Neuroimaging adjuncts to MRI: SHARON E. BYRD, MD aFall 1998 OVER THE PAST several years, three simple software packages have been developed to provide additional information to routine magnetic resonance (MR) images to further aid in the evaluation and interpretation of abnormalities affecting the central nervous system (CNS) in newborns, infants, and children. These software packages consist of magnetic resonance angiography (MRA), which assesses the intra- and extra-cranial circulation (arteries and veins); magnetic resonance spectroscopy (MRS), which assesses metabolic activity of the brain or brain abnormalities; and cinè MR of cerebrospinal fluid (CSF) flow and analysis, which assesses movement of CSF as well as brain and spinal cord pulsations. These software packages can easily be added on to any high field strength MR scanner. The physics of these adjuncts is based on routine spin echo or gradient echo imaging. All of them–MRA/MRS/Cinè MR of CSF flow and analysis—must be performed in conjunction with a routine MRI exam. While the MRI exam provides the anatomical picture of the CNS and may demonstrate the abnormality, these adjuncts provide further information about the CNS and more precisely delineate the abnormality; furthermore, they can be performed in 10 to 30 minutes of additional scan time. The only contraindications to performing these adjuncts are the routine contraindications to an MRI exam such as a pacemaker. In infants and young children in whom sedation is necessary for the MRI exam, the usual sedation dose is more than adequate for completion of the MRI along with the MR adjunct pulse sequences. Magnetic Resonance Angiography (MRA) MRA is a new technique that uses gradient echo pulse sequences and computer post-processing technology to create a vascular flow image. It is an extremely useful adjunct to traditional imaging modalities as a screening device for vascular disease. Two MRA pulse sequences—time of flight (TOF) and phase contrast (PC)—are used singularly or in combination to evaluate a host of vascular abnormalities such as vascular malformations (true arteriovenous malformations [AVM] including vein of Galen aneurysm, venous angiomas, dural and pial AVMs), aneurysms, vascular occlusions (arterial obstruction or stenosis), arterial dissection, arteritis, and venous thrombosis.   FIGURE 1A. MRI (T1 axial): 18-day-old with hemorrhagic infarction in left posterior temporo-occipital area (arrows). FIGURE 1B. MRA (2D PC sagittal): Neonate in figure 1a with hemorrhagic infarction secondary to venous thrombosis with occlusions of both transverse sinuses (large and small arrows). The role of MRA is as follows: It is the best modality to evaluate venous thrombosis (Figures 1a, 1b). It is excellent for demonstrating giant intracranial aneurysms (2.5 cm or greater) and can demonstrate aneurysms as small as 3 to 5 mm (Figure 2).  FIGURE 2. MRA (3D TOF coronal): Child with severe headaches and meningismus with berry aneurysm of supraclinoid bifurcation of right internal carotid artery (arrow). Therefore, it can be used as a screen in children where there is a predisposition to the development of aneurysms (such as in children with coarctation of the aorta or polycystic kidneys). It can screen for AVMs and be used to follow AVMs before and after treatment (Figure 3).  FIGURE 3. MRA (3D PC sagittal): Newborn with hydrocephalus secondary to arteriovenous fistula from carotid and vertebral arterial circulation to a vein of Galen aneurysm (arrow). And it helps to provide additional information to diagnose arterial dissection, occlusion, and/or stenosis (Figures 4a, b).  FIGURE 4A. MRA (3D TOF sagittal): Normal arterial intracranial circulation.  FIGURE 4B. MRA (3D TOF sagittal): Child with headaches and hemiparesis demonstrating occlusion of the major branches of the carotid intracranial circulation with central deep collateral arteries [moya moya-puff of smoke (arrows)]. Although MRA is an excellent screening device for arterial disease, it is not as sensitive as conventional angiography. Conventional angiography is still the "gold standard" for evaluating arterial disease and abnormalities such as aneurysms, AVMs and arterial occlusions, or stenoses. These abnormalities must still be evaluated with conventional angiography even when demonstrated on the prescreen MRA because subtle abnormalities may be missed on the MRA. But the MRA as a screen can help determine which children need to be further evaluated. Therefore it can reduce the overall need for conventional angiography.1—3 Magnetic Resonance Spectroscopy (MRS) MRS provides a simple means to assess the metabolic activity of the brain and/or brain abnormalities. One of the largest barriers to its clinical use has been the misconception that only a physicist can produce reliable spectra. The two most common types of MR spectroscopy are based on the resonance of phosphorus (P) or the hydrogen (H) proton. The majority of proton MRS techniques for neurospectroscopy are easily implemented on existing commercial MR scanners with only minor software modifications. Neurospectroscopy can be performed with the same equipment and staff that are used for routine MRI exams. Proton MRS is more readily performed and interpreted than phosphorus MRS. We perform proton MRS at Children's Memorial Hospital using a single voxel (localizer) for qualitative and quantitative MRS studies. Localized single-voxel proton MRS consists of analyzing the H proton resonance of a voxel (cubic volume) of a specific local area of the brain (Figure 5a).  FIGURE 5A. MRI (T2 axial): Normal MRI in 14-month-old with slight delay in development with localizer voxel over right posterior temporo-parietal white matter for spectroscopic analysis. The size of the cube can vary from 1 to 8 cm3, although most voxels are from 1.5 to 3 cm3 . Any localized area of brain tissue can be analyzed. Qualitative MRS studies consist of producing a spectrum (a graph) of the metabolites within the chosen voxel (cube) of brain tissue, while quantitative studies measure the actual amounts or ratios of the metabolites within the chosen voxel (Figure 5b). Quantitative analysis (quantification) is performed by feeding the raw data of the spectrum into a SUN computer to produce the values for specific brain metabolites.  FIGURE 5B. MRS (Steam qualitative spectrum of voxel in Figure 5a) demonstrating slight decrease in NAA peak indicating a delay in neuronal maturation. MRS provides a means to assess the biochemical characteristics of brain cells and the cells of brain pathology through direct, noninvasive assay of cerebral metabolites. The most common metabolites within normal brain tissue consist of N-acetylasparate (NAA), creatine (CR), choline (CHO), glutamate-glutamine (GLX), and myo-inositol (MI). Lactate (LAC) and lipids are usually abnormal brain metabolites. NAA is a neuronal marker and is only present in neurons. CHO is involved with cell membrane synthesis. CR is involved in the high energy transport system (it helps to maintain mitochondrial energetics). GLX is involved as a neurotransmitter. LAC is involved in anaerobic glycolysis and can be seen in malignant brain tumors, infarction, and necrosis. Lipid can be seen in breakdown of myelin and brain tumors. There are two pulse sequences with H MRS-stimulated echo acquisition mode (STEAM), which is T1-weighted, and point-resolved spectroscopy (PRESS), which is T2-weighted. These two pulse sequences allow for full MRS evaluation of the voxel of brain tissue. The MRS spectrum is along a relative scale established in parts-per-million (ppm) for peak positions of the resonance of these common brain metabolites (see Table 1). The scale is from 0 to 4 ppm, and the scale is read from right to left with 0 on the right and 4 on the left (Figure 5b and Figure 6).  FIGURE 6. MRS (PRESS qualitative spectrum): 5-month-old with slight delay in development–voxel over white matter demonstrating decreased NAA peak indicating delay in neuronal maturation.
Because it is more sensitive than MRI in demonstrating a delay in development of the brain (Figure 5b and Figure 6, MRS is useful in following the normal maturation of the brain in infants and children and can be used along with the MRI exam to follow the myelination process. MRS can also characterize a brain mass on MRI as a neoplasm (Figure 7) and therefore has been used extensively in evaluating children with these conditions.  FIGURE 7A. MRI (T2 axial): Child with tuberous sclerosis with voxel (arrow) over mass in frontal horn of left lateral ventricle.  FIGURE 7B. MRS (STEAM): Same child as Figure 7a with markedly decreased NAA and markedly elevated CHO, which indicates mass in neoplasm. Furthermore, it is helpful in following the treatment response of brain neoplasms and can differentiate radiation necrosis from a residual or recurrent neoplasm (Figure 8). In fact, within two weeks of initial treatment for brain neoplasm with radiation or chemotherapy, MRS can determine the neoplasm's response. Whereas with routine MRI exams, it may take 3 to 6 months to determine if a brain tumor is responding to treatment, MRS can determine if residual tumor is inactive (in remission) or active.  FIGURE 8A. MRI (T2 axial): Child with suprasellar neoplasm treated with surgery, radiation and chemotherapy with voxel (arrow) for MRS placed over residual mass.  FIGURE 8B. MRS (PRESS): Same child as Figure 8a demonstrating a noisy spectrum with no metabolites indicating radiation necrosis and no residual neoplasm. Children who have evidence of brain dysfunction such as speech and language disorders, intellectual or motor deterioration, mental retardation and/or psychiatric disorders, or attention deficit disorders commonly have normal MRI exams. But MRS demonstrates abnormal brain functioning. MRS can also be used to help diagnose and follow metabolic and neurodegenerative brain disorders. It can help diagnose and follow the extent of brain damage in perinatal asphyxia (hypoxic-ischemic encephalopathy) and head trauma from accidental and non-accidental causes (such as child abuse).4–8 Cinè MR of CSF flow and analysis Cinè MR of CSF flow is a technique using cardiac gated gradient echo phase contrast MR imaging in which acquired data is displayed qualitatively as CSF flow images in a closed-loop cinè format. The data can also be displayed quantitatively as quantification analysis values of flow velocity and volume flow rate in a graph with numerical calculations.  MRI (T2 sagittal): Child with Chiari I (arrow). The most commonly used format is to display the data qualitatively as speed, magnitude, and/or phase-reconstructed images with actual visualization of CSF flow on cinè (Figures 9, 10a, b). Cinè MR demonstrates the pulsatile motion and not the slow bulk flow of CSF. The phase images are the best for the demonstration of the pulsatile flow of CSF on cinè. Pulsatile flow of CSF occurs as a result of expansion of the brain during systole and relaxation during diastole. Therefore, pulsatile flow is bidirectional with CSF flowing caudally during systole and cranially during diastole.  FIGURES 10A, B. Cinè MR CSF flow (phase images): Same child as Figure 9 with CSF in anterior basal cisterns and upper cervical as hyperintense (white-arrows) and as hypointense (black-arrows). Obstruction to CSF flow posterior aspect of posterior fossa and posterior foramen magnum (arrow heads) due to the Chiari I. The signal intensities of normal CSF flow patterns on cinè MR demonstrate CSF flow as hyperintense (white) during systole (caudal downward flow) and hypointense (black) during diastole (cranial upward flow) through a cycle Figures 10a, b). Cinè MR CSF flow studies are usually performed in the sagittal projection (although coronal and axial planes can also be used) with a scan time of 3 to 4 minutes for each pulse sequence. A normal study demonstrates hyperintense–hypointense CSF in structures such as the ventricles, aqueduct of Sylvius, foramen of Magendie, basal cisterns of the brain, and ventral and dorsal spinal subarachnoid spaces. The lack of demonstration of flow (hyperintensity and hypointensity) of CSF is a sign of obstruction of a CSF pathway (Figures 10a, b). CSF pulsatile flow normally diminishes as it proceeds caudally down the spinal canal. Therefore, the signal on the cinè MR CSF flow images will decrease in the lumbosacral region with decreased resolution compared to the cranial and cervical areas. Cinè MR CSF flow studies are indicated whenever it is necessary to evaluate the flow of CSF, rule out an obstruction in the CSF pathway, evaluate CSF flow around Chiari I and II malformations (Figures 9, 10a, b), evaluate spinal canal stenosis, evaluate postoperative decompressive procedures on the spine and brain, and evaluate third ventriculostomy.9–10 Conclusion By adding these adjunct pulse sequences—MRA, MRS and cinè MR CSF flow studies to the routine MRI exam-better delineation and more precise diagnosis of abnormalities that affect the pediatric CNS can be obtained. REFERENCES 1. Anderson CM, Edelman RR, Turski PA: Clinical Magnetic Resonance Angiography. New York City: Raven Press, 1993. 2. Potchen EJ, Haacke EM, Siebert JE, Gottschalk A: Magnetic Resonance Angiography: Concepts and Applications. St. Louis: Mosby, 1993. 3. Keller PJ, Drayer BP, Fram EK: Neuroimaging Clinics of North America. Philadelphia: W. B. Saunders Co., 1992. 4. Orrison, Jr. WW, Lewine JD, Sanders JA, Hartshorne MF: Functional Brain Imaging. (1st ed.) St. Louis: Mosby, 1995. 5. Ross BD, Moats RA, Mandigo JC, Michaells T: Neurospectroscopy is ready for prime time. MR, the Newsmagazine of Magnetic Resonance 1994:20–24. 6. Ross B, Michaells T: Clinical applications of magnetic resonance spectroscopy. Magn Reson Q 1994;10:191–247. 7. Izika AA, Vigneron DB, Ball WS, Dunn RS, Kirks DR: Localized proton MR spectroscopy of the brain in children. Magn Reson Imaging 1993;3:719–729. 8. Izika AA, Ball, Jr. WS, Vigneron DB, Dunn RS, Kirks DR: Clinical proton MR spectroscopy of neurodegenerative disease in childhood. Am J Neuroradiol 1993:14:1267–1281. 9. Enzmann DR, Pelc NJ: Cerebrospinal fluid flow measured by phase-contrast cine MR. Am J Neurol 1993, 14:1301–07. 10. Quencer RM, Donovan Post MJ, Hinks RS: Cinè MR in the evaluation of normal and abnormal CSF flow: Intracranial and intraspinal studies. Neuroradiology 1990, 32:371–391. |
||||||||||||||||||||||||||||||||||