| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Children's Hospital of Philadelphia, 34th and Civic Center Blvd., Philadelphia, PA 19154.
Received August 23, 1999;
accepted after revision February 22, 2000.
Address correspondence to A. M. Hubbard.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A retrospective review found 66 MR imaging studies of 63 fetuses, 16-39 weeks' gestation age (mean, 25 weeks' gestation), who were referred between June 1996 and May 1999 for evaluation of non-central nervous system anomalies revealed on prenatal sonography. All fetuses had normal brains and spines on prenatal sonography. The standard MR imaging protocol included axial, sagittal, and coronal half-Fourier acquisition single-shot turbo spin echo (HASTE); sagittal and coronal two-dimensional fast low-angle shot (FLASH); and axial turbo T1-weighted FLASH images through the fetal brain. Structures that we analyzed were the fourth ventricle, the cisterna magna, the vermis, the cerebellar hemispheres, and the brainstem. Using the HASTE sequences, we documented gestational age-specific signal intensity changes in the cerebellar hemispheres and the brainstem.
RESULTS. The posterior fossa anatomy was sufficiently well defined to exclude abnormalities of the fourth ventricle and cerebellar vermis in all cases. Because of high T2-weighting, good contrast enhancement, and good signal-to-noise ratios, HASTE images provided the best anatomic definition of the posterior fossa.
CONCLUSION. Normal posterior fossa anatomy can be adequately shown on ultrafast MR images, which can be helpful when prenatal sonography is equivocal.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Prenatal sonography in the second trimester is the technique of choice for the screening of brain anomalies in utero. Sonography provides real-time imaging that is safe, readily available, noninvasive, and cost-effective [3, 4]. However, prenatal sonography has its limitations, particularly in the examination of the posterior fossa, which is often difficult given the compact anatomy. To evaluate the posterior fossa sonographically, a transthalamic axial view of the head angled posteriorly toward the cerebellum is obtained. From this one view, the cerebellum, cerebellar vermis, cisterna magna, and nuchal soft tissues are evaluated [5]. However, the degree of posterior angulation has not been standardized because of lack of a single anatomic landmark. Therefore, depending on the degree of angulation, inferior vermian defects and other subtle posterior fossa anomalies can go undetected. Conversely, the cisterna magna can appear falsely enlarged if the angle of inclination of the transducer is increased [6, 7]. In addition, evaluation of the posterior fossa in fetuses older than 33 weeks' gestation can be difficult because of interference of beam penetration with the ossified cranium or because the head is engaged in the maternal pelvis [3]. Fetal sonography can also be limited by other factors such as maternal obesity, gasfilled maternal structures, oligohydramnios, and fetal motion [4, 8, 9].
Fetal MR imaging is helpful in the evaluation of brain anomalies detected on sonography and in cases in which sonography was equivocal or inadequate [1, 3, 8, 10,11,12,13,14,15,16,17]. Earlier MR imaging studies in the mid 1980s to early 1990s were limited by fetal motion, so maternal sedation with diazepam or fetal paralysis with pancuronium or vecuronium bromide was used to obtain quality images, introducing risks to the mother and fetus [1, 3, 10, 12, 14, 16, 18]. Because of the development of ultrafast MR imaging in the 1990s, quality multiplanar images can now be obtained in milliseconds, essentially eliminating the problem of motion artifacts and obviating premedication.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Informed consent was obtained for all patients. MR imaging was performed on a 1.5-T Vision Magnetom System (Siemens Medical Systems, Iselin, NJ) with a body phased array coil. Our standard protocol included half-Fourier acquisition single-shot turbo spin-echo (HASTE), two-dimensional fast low-angle shot (2D FLASH,) and turbo T1-weighted FLASH images. HASTE images (TR/TE, 4.4/64 msec; flip angle, 120°; matrix, 128 x 256; slice thickness, 6 mm with no gap) were obtained through the fetus in the axial, sagittal, and coronal planes. Two-dimensional FLASH images (174/4.1 msec; flip angle, 80°; matrix, 128 x 256; slice thickness, 5 mm with a 0.2-mm gap) were obtained in the sagittal and coronal planes. Turbo T1 FLASH images (11.0/5.3; flip angle, 15°; matrix, 128 x 256; slice thickness, 5 mm with a 0.2-mm gap) were obtained from the fetal brain through the pelvis in the axial plane. Whenever possible, patients were asked to hold their breath on all sequences with a scan time equal to or less than 22 sec.
All images were reviewed by three radiologists: two senior attending radiologists and the first-year pediatric radiology fellow. To determine which sequence revealed posterior fossa anatomy best, each of the three sequences was evaluated in terms of contrast and resolution. The anatomic structures analyzed were the cerebellar hemispheres, the vermis, the cisterna magna, the fourth ventricle, and the brainstem. Any signal intensity changes of these structures seen with fetal maturation were documented.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The standard sequences we used displayed enough posterior fossa anatomy to exclude morphologic anomalies in all cases (Figs. 1 and 2). We found that 5- and 6-mm-thick slices were adequate for excluding anomalies of the vermis and cisterna magna. The thickness of the slices was the thinnest possible for the sequences used. The HASTE sequence was sufficient in 64 cases. In two cases, however, the 2D FLASH sequence was used to confirm the anatomy because the appropriate planes were not performed on the HASTE sequences. In addition, characteristic signal intensity changes on HASTE sequences were detected in the cerebellum and brainstem (Table 1). Between 16 and 19 weeks' gestation, the cerebellar hemispheres and most of the brainstem showed homogenous, intermediate signal intensity. A lobulated area of low signal intensity was seen in the dorsal aspect of the midbrain as early as 16 weeks' gestation and certainly by 20 weeks' gestation in all cases. This area corresponded anatomically to the tectum (Fig. 3). By 20-23 weeks' gestation, lower signal intensity that corresponded anatomically to the cerebellar cortex was seen in the periphery of the cerebellar hemispheres. At the same gestational age, a lower signal-intensity vertical band was seen in the posterior aspect of the pons and the medulla (Fig. 3); the band reached the midbrain by 32 weeks' gestation (Fig. 4A,4B). This band corresponded to the region of the medial longitudinal fasciculus. No other changes in signal intensity were noted in the brainstem after 32 weeks' gestation.
|
|
|
|
|
|
By 23-26 weeks' gestation, minimal undulation of the periphery of the cerebellar hemispheres was seen, and by 32 weeks' gestation prominent folia were identified (Fig. 4A,4B) that increased in number as the fetus approached term. The folial pattern was best depicted on the axial images as a layered pattern of alternating intermediately low- and high-signal-intensity bands (Fig. 5). By 27 weeks' gestation, the cerebellar tonsils were seen. In addition, by 26-27 weeks' gestation, three distinct cerebellar layers were identified that were best detected on the coronal views. The outermost thin layer of low signal intensity corresponded anatomically to the cerebellar cortex, and the thicker middle layer of high signal intensity corresponded to the cerebellar white matter. The deepest layer was a thin layer of low signal intensity that outlined the fourth ventricle on either side and corresponded to the dentate nucleus (Fig. 6). The middle cerebellar peduncles were first appreciated at 24 weeks' gestation as bilateral bands of low signal intensity extending from the pons to the cerebellar hemispheres on the axial images (Fig. 7).
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Multiple case reports and studies have proven the usefulness of MR imaging in defining fetal anatomy and pathology [3, 4, 8,9,10, 12,13,14, 16, 17, 19, 20]. MR imaging provides superior resolution and soft-tissue characterization compared with sonography. It is more efficient in the evaluation of the subarachnoid spaces, ventricular walls, and intraparenchymal architecture [1, 3, 9, 11, 19]. In addition, because images can be obtained in the sagittal and coronal planes, MR imaging is superior to sonography in the examination of the posterior fossa for anomalies such as Dandy-Walker malformation and Arnold-Chiari malformations. However, most fetal MR imaging studies to date have concentrated on supratentorial brain anatomy, with particular attention to neuronal migration patterns. Minimal focus has been placed on normal anatomy and development of the posterior fossa on MR imaging [1, 3, 8, 11, 13, 16, 17, 21].
As stated earlier, from previous studies we learned that MR imaging templates of fetal cerebral development can lag behind their gestational age—specific histologic correlates [1, 2]. A recent study by Chong et al. [2] analyzed normal cerebellar development on MR imaging on 26 fetal specimens ranging from 9 to 24 weeks' gestation. Comparison was made to age-specific histologic specimens. They found a lag of up to 5 weeks between the appearance of some structures histologically and their visualization on MR imaging. They attributed this discrepancy to the limits of resolution of MR imaging, particularly in the first and early second trimesters, when structures are small [2]. Therefore, it would be erroneous to interpret MR imaging findings of the posterior fossa on the basis of the expected histologic findings for a given gestational age. An in vivo MR imaging template of normal fetal posterior fossa anatomy and development is essential for proper interpretation of prenatal MR imaging of the brain.
Most previous MR imaging studies of the development of the supratentorial brain in fetuses have documented normal anatomy using T1-weighted sequences [1, 11, 21]. However, several reports have shown the HASTE sequence to be excellent for examining the developing fetal brain [9, 10, 14, 15]. Our study agrees with these previous reports. The HASTE sequence is an ultrafast, heavily T2-weighted sequence with good signal-to-noise ratio. Because each image is acquired sequentially in less than 1 sec, motion artifacts are fewer than on the 2D FLASH and turbo T1 FLASH sequences [9, 10]. Moreover, if motion occurs during the acquisition, only the images acquired during motion are affected [20], which results in better effective resolution. This is not the case with the two other sequences, in which all images would be affected by fetal motion occurring at some point during the acquisition. The fetal brain has a lower protein content and higher water content than the normal adult brain, with an extracellular space equaling 40% of the total brain versus 20% in the adult. This higher water content results in long T2 relaxation characteristics in the fetal brain. Therefore, subtle differences in signal intensity in the brain parenchyma are best shown using a heavily T2-weighted sequence like HASTE to highlight T2 differences [1, 19]. This technique results in better contrast. The 2D FLASH and turbo T1 FLASH sequences were T1-weighted and therefore were not optimal for defining fine detail in the fetal brain.
Changes of signal intensity in the brainstem and cerebellum were not evident on MR imaging until 20-23 weeks' gestation, which correlates with recent findings of Levine et al. [20], who analyzed normal fetal anatomy using a similar ultrafast sequence and found that the resolution of MR imaging was limited in live fetuses of fewer than 20 weeks' gestation, probably because of the small size of the structures being analyzed. We agree with Girard et al. [1] that signal intensity changes (high on T1-weighted and low on T2-weighted images) in the cerebellar cortex, dentate nucleus, and tectum are the result of relatively high cellularity in these areas composed of gray matter, which contains less water than the adjacent white matter. This assumption is based on histologic correlates that show hypercellularity in these areas [22]. In addition, we also agree with Girard et al. that the signal intensity change of the posterior aspects of the pons and medulla (high on T1-weighted and low on T2-weighted images) corresponds to the myelination of the medial longitudinal fasciculus, which is also first seen histologically at 20 weeks' gestation.
The 1997 in vitro study of posterior fossa development by Chong et al. [2] first showed changes in cerebellar signal intensity at a similar time (19 weeks' gestation). However, those researchers first noted signal intensity changes in the posterior brainstem at 17 weeks' gestation, which is before the myelination of the medial longitudinal fasciculus has occurred. They postulated that these signal intensity changes in the dorsal brainstem are not the result of myelination, but instead result from differences in water content between densely packed gray matter in the pontine tegmentum (dorsal pons) and the nonmyelinated tracts with scattered nuclei seen in the basis pontis. A factor to be taken into consideration is that they used only T1-weighted imaging. In addition, they studied formalin-fixed specimens. The in vivo study by Girard et al. [1] also used T1-weighted images and first detected signal intensity changes in the brainstem at 23 weeks' gestation. Therefore, the HASTE images were comparable to, if not more sensitive than, the T1-weighted sequences in terms of detecting subtle signal intensity differences in vivo.
Unlike previous MR imaging studies of fetal development of the supratentorial brain that showed a five-layered pattern of neuronal migration correlating to the expected histologic pattern, in our study neuronal migration in the cerebellar hemispheres was not evident on MR imaging [1, 11, 21]. This was also the case in the recent in vitro study by Chong et al. [2]. However, in our study, a three-layered pattern was noted in the cerebellar hemispheres by 26-27 weeks' gestation, which corresponds to the development of the cerebellar cortex, cerebellar white matter, and dentate nucleus (Fig. 6). This pattern is based on coronal fetal histologic sections that show these three basic layers composing the cerebellar hemispheres. Two of these layers, the cerebellar cortex and the dentate nucleus, are hypercellular and therefore correspond to their low signal intensity on the HASTE sequences [23]. This three-layered pattern was also seen in the recent in vitro study by Chong et al. [2], although it was seen earlier, at 19 weeks' gestation. The same recent study by Chong et al. identified folia in the vermis at 16 weeks' gestation. Histologically, folia are first identified at 12-13 weeks' gestation [24, 25]. Our study first identified folia at 23-26 weeks' gestation, yielding a longer lag time between the appearance of folia histologically and their appearance on MR imaging. The discrepancy between the two studies may lie in the fact that Chong et al. used formalin-fixed specimens, which yielded better resolution. Our study included only eight fetuses between 16 and 19 weeks' gestation. However, we believe that we were not able to identify folia before 23 weeks' gestation because of the limitation of resolution of MR imaging in the early second trimester [2].
In conclusion, although MR imaging will not replace sonography in the evaluation of the fetal brain, it plays a crucial role as an adjuvant study, especially in the evaluation of the posterior fossa. As the fetus matures, predictable changes in signal intensity can be seen that correspond to anatomic and histologic differentiation. Using the HASTE sequence, we found what we believe to be the first template of normal fetal posterior fossa development revealed on MR imaging.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Limperopoulos, R. L. Robertson Jr., O. S. Khwaja, C. D. Robson, J. A. Estroff, C. Barnewolt, D. Levine, D. Morash, L. Nemes, L. Zaccagnini, et al. How Accurately Does Current Fetal Imaging Identify Posterior Fossa Anomalies? Am. J. Roentgenol., June 1, 2008; 190(6): 1637 - 1643. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O.A. Glenn and A.J. Barkovich Magnetic resonance imaging of the fetal brain and spine: an increasingly important tool in prenatal diagnosis, part 1. AJNR Am. J. Neuroradiol., September 1, 2006; 27(8): 1604 - 1611. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Chen, E. M. Simon, J. C. Haselgrove, L. T. Bilaniuk, L. N. Sutton, M. P. Johnson, D. M. Shera, and R. A. Zimmerman Fetal Posterior Fossa Volume: Assessment with MR Imaging Radiology, March 1, 2006; 238(3): 997 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. F. Reichel, R. M. Ramus, J. T. Caire, L. S. Hynan, K. P. Magee, and D. M. Twickler Fetal Central Nervous System Biometry on MR Imaging Am. J. Roentgenol., April 1, 2003; 180(4): 1155 - 1158. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
