| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University Hospital
Zürich, Rämistr.
100, CH-8091 Zürich, Switzerland.
2
Department of Obstetrics, University Hospital
Zürich, CH-8091
Zürich, Switzerland.
3
Division of Biostatistics, University of Zürich,
Sumatrastr. 30, CH-8006 Zürich,
Switzerland.
4
Department of Pathology, University Hospital
Zürich, CH-8091
Zürich, Switzerland.
Received October 5, 1999;
accepted after revision November 16, 1999.
Address correspondence to R. A. Kubik-Huch.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Thirty women with complicated pregnancies (mean age of gestation, 190 ± 54 days) underwent T2-weighted ultrafast MR imaging. MR images were analyzed with regard to diagnostic confidence in assessing abnormalities of fetal organ systems, and data were correlated with postpartum findings or necropsy. Results were compared with those of prenatal sonography.
RESULTS. Using receiver operating characteristic curve analysis, diagnostic confidence of MR imaging was best for assessing the brain (area under the curve [Az] = 0.96) and spinal canal (Az = 1.0), uteroplacental unit (Az = 0.93), and lungs (Az = 0.91). Results for the heart (Az = 0.63) and extremities (Az = 0.77) were significantly lower than that of other organs (p < 0.001). Diagnostic accuracy increased with gestational age. No statistically significant difference between sonography and MR imaging was found for the detection of abnormality in any organ system. In three fetuses, MR imaging was superior to sonography in characterizing cerebral abnormalities. MR imaging was inferior to sonography in characterizing abnormalities of the heart and extremities.
CONCLUSION. Our results indicate that ultrafast MR imaging can be used for in vivo fetal imaging, especially in assessing cerebral abnormalities. However, MR imaging should be restricted to situations in which sonographic findings are ambiguous or impaired.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Advantages inherent to MR imaging include unsurpassed soft-tissue contrast enhancement and multiplanar imaging capabilities. Similar to sonography, MR imaging does not expose patients to ionizing radiation, and no clinical or experimental evidence of teratogenic or other adverse fetal effects of MR imaging in pregnancy exists [1,2,3,4,5,6]. Therefore, MR imaging is suited for maternal imaging during pregnancy and (as an operator-independent technique) as an adjunct to sonography in assessing complex prenatal abnormalities.
Until recently, the clinical use of in utero fetal MR imaging was limited by marked interference from motion artifacts caused by long acquisition times. Even with indirect fetal sedation via IV sedation of the mother or direct fetal sedation via cordocentesis, image quality of conventional spin-echo sequences remained insufficient to permit the delineation of small fetal structures. Nevertheless, the potential of MR imaging was revealed with studies of cerebral abnormalities, such as midline or posterior fossa deformities [7,8,9,10,11].
Technical advances in ultrafast MR imaging, particularly gradient-echo sequences with reduced TRs and TEs or echoplanar sequences, have led to further attempts to apply MR imaging to prenatal diagnoses [12,13,14,15,16,17]. Limited signal-to-noise ratios, susceptibility artifacts, and poor spatial resolution impaired the implementation of these techniques in clinical protocols.
Recently, the ability of rapid relaxation contrast-enhanced MR imaging to depict normal fetal anatomy and abnormality has been reported [18,19,20,21]. We examined the capability of ultrafast single-shot fast spin-echo imaging to assess different fetal organ systems compared with prenatal sonography, using autopsy or postpartum imaging as a standard of reference.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In 25 pregnant women, MR imaging was performed because of abnormal findings on prenatal sonography, amniocentesis, or both. In one patient, an abnormal position of the fetal head was observed during routine MR pelvimetry, prompting fetal MR imaging. Four women underwent fetal MR imaging because of suspected maternal disease. The mean age of gestation at the time of MR imaging was 190 ± 54 days (range, 75-278 days).
Before MR imaging, prenatal sonography was performed in all patients by the same investigator, who was experienced in the setting of high-risk pregnancies. Sonography was performed using a 4-MHz sector probe (128xP/10; Acuson, Mountain View, CA) or a multifrequency (3.5-5.1 MHz) probe (Elegra; Siemens, Issaquah, WA).
MR imaging was performed on a 1.5-T system (Signa Echo Speed; General Electric Medical Systems, Milwaukee, WI). After a localizing gradient-echo sequence, ultrafast T2-weighted single-shot fast spin-echo MR images (TR/TE, infinite/90; bandwidth, 32 kHz; field of view, 16 x 28 cm; matrix, 256 x 192; slice thickness, 3-5 mm; gap, 1.5 mm; number of excitations, 0.5) were collected according to fetal position in the axial, coronal, and sagittal planes, perpendicular or parallel to the specific fetal body part being studied. The acquisition time for each individual image was 0.8 sec. A torso or pelvic phased array coil was used depending on fetal size and maternal abdominal circumference. In patients referred for the evaluation of maternal disease, additional T1-weighted spin-echo and T2-weighted single-shot fast spin-echo MR sequences were performed, using a larger field of view that encompassed the entire maternal abdomen.
Full autopsies (n = 10) were performed by the same experienced pediatric pathologist of all aborted and diseased infants, except for four fetuses in whom autopsy was rejected by the parents. Autopsies included a comprehensive description and photographic record of the whole fetus and dysmorphic features, including supplementary conventional radiography and tissue specimens for microscopy, cytogenetics, or metabolic studies (as appropriate).
MR images were prospectively analyzed by an experienced abdominal radiologist and a pediatric neuroradiologist in consensus with regard to fetal and maternal normal structures and abnormalities, including the placenta, uterus, and umbilical cord. The MR examinations were assessed independently of and blinded to sonographic results and standards of reference.
Each fetal organ system (brain, spinal canal, heart and great vessels, lung, liver and spleen, urinary tract, facial structures, and extremities) and the uteroplacental unit were assessed on a five-point scale (1 = normal, 2 = probably normal, 3 = inconclusive, 4 = probably abnormal, 5 = abnormal). Those data were compared with the standard of reference. For the standard of reference, pathologic findings were those proven by either clinical follow-up or necropsy. Normal findings were those that proved normal at pathology or had no suspicion of abnormality on any imaging technique or clinical follow-up.
The final diagnoses were congenital multiple arthogryposis, brain
infarction, chromosomal anomaly (triploidy 69, tetrasomy 12p, trisomy 13,
triploidy 69, XXY), Chiari II deformity (n = 3), lobar
holoprosencephaly, ectopic intrauterine pregnancy,
müllerian duct aplasia-renal
aplasia-cervicothoracic somite dysplasia-association with hydrocephalus,
arachnoid cyst (n = 2), anencephaly, intracerebral cystic mass
lesion, indeterminate dysmorphic syndrome, congenital fibrosarcoma, skeletal
dysplasia, aquaeductus stenosis, immature cervical carcinoma, and occipital
cele. Additional findings of the uteroplacental unit or the maternal body
included septate uterus (n = 2), myomatous uterus, maternal cervical
carcinoma, and ovarian mucinous tumor. MR imaging and postpartum findings were
normal in two fetuses with sonographically suspected ventriculomegaly and in
one patient with an elevated 
-fetoprotein level.
Thirteen fetuses were male and 14 were female. One fetus presented with a female phenotype but was a male. In one ectopic intrauterine pregnancy and in one fetus with a complex malformation, the gender was unknown.
Spontaneous abortion occurred in one patient with an ectopic pregnancy. Labor was induced in two patients because of intrauterine fetal death. The pregnancies of eight women were terminated. The other women delivered spontaneously or by cesarean section.
For statistical analysis, diagnostic confidence scores of the individual organ systems were compared using Wilcoxon's signed rank test with Bonferroni adjustment. The score for each organ system was then correlated with the gestational age at the time of MR imaging using Spearman's rank correlation.
Receiver operating characteristic curves were determined for each organ system except the liver and the spleen because we had no pathologic findings for these organs.
MR imaging findings were compared with sonography findings using the McNemar test. For this purpose, MR imaging scores were categorized as either normal (scores 1-3) or abnormal (scores 4 and 5).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In comparing the different organ systems, the diagnostic confidence score in assessing the heart was significantly lower than that of the brain, liver and spleen, urinary tract, spinal canal, and uterus and placenta (p < 0.001). The scores for the extremities were significantly lower than those of the spinal canal and the uteroplacental unit (p < 0.001).
Using receiver operating characteristic curves, interpreters were in best agreement for the assessment of the brain (Az = 0.96) and spinal canal (Az = 1.0), followed by the uteroplacental unit (Az = 0.93) and lungs (Az = 0.91). Agreement was mediocre for the urinary tract (Az = 0.79), facial structures (Az = 0.83), extremities (Az = 0.77), and heart (Az = 0.63) (Fig. 1A,1B).
|
|
Diagnostic confidence scores significantly increased with gestational age at the time of MR imaging for the lung (R = 0.54; p = 0.002) and spinal canal (R = 0.52, p = 0.003). Although the results were statistically insignificant after the Bonferroni adjustment, the same trend was observed for the facial structures (R = 0.42, p = 0.02), liver and spleen (R = 0.48, p = 0.007), and urinary tract (R = 0.44, p = 0.019).
MR imaging performed similarly to sonography in detecting abnormalities (Table 1 and Fig. 2A,2B,2C), and the McNemar test revealed no statistically significant difference between the two imaging techniques.
|
|
|
|
MR imaging was slightly superior to sonography with regard to cerebral abnormalities only. Advantages of MR imaging included the better delineation of the cerebral cortex and improved assessment of the myelination and gyration. In three fetuses, MR imaging corrected previous sonographic diagnoses: in the first fetus, sonography findings were suggestive of Dandy-Walker deformity and MR imaging revealed a normal fourth ventricle and provided the correct diagnosis of a posterior fossa arachnoid cyst (Fig. 3A,3B,3C); in the second fetus with müllerian duct aplasia-renal aplasia-cervicothoracic somite dysplasia-association, sonography provided a false diagnosis of Chiari II malformation, whereas MR imaging revealed findings suggestive of supratentorial hydrocephalus and normal insertion of the tentorium. However, sonography of the second fetus also revealed the cervical vertebrae malformation and an absent right umbilical artery, indicating a complex malformation and prompting the parental decision to terminate. These findings were overlooked on MR imaging. In the third fetus, sonography revealed normal findings, but previously unknown watershed infarctions were diagnosed at 252 days' gestation (probably unrelated to the abnormal fetal position) and confirmed on postpartum cranial sonography and MR imaging (Fig. 4A,4B,4C).
|
|
|
|
|
|
MR imaging was inferior to sonography in assessing the heart because of missing flow information, otherwise provided on Doppler sonography. In this study, only two cardiac deformities were detected. In one fetus with a complex dysmorphic syndrome, necropsy revealed tetralogy of Fallot (a ventricular septal defect was revealed on sonography) and MR imaging revealed an enlarged heart. In the other fetus with chromosomal abnormalities (69, XXY [30]), sonography revealed a ventricular septal defect with an overriding aorta. Again, MR imaging revealed only an enlarged heart.
MR imaging was also limited in the assessment of extremities. Pathologic contractions of the extremities were revealed on MR imaging in a fetus with complex dysmorphic syndrome; however, syndactylia seen on sonography was overlooked. In the fetus with fibrosarcoma of the thigh, MR imaging revealed the tumor, but sonography was superior in depicting the destruction of the femur (Fig. 5A,5B,5C).
|
|
|
All fetuses with meningomyeloceles and Chiari II deformity presented with normal lower extremities and without clump feet on MR images. However, normal movement was revealed only on sonography as a real-time technique. Only sonography revealed missing movement of the extremities in the fetus with Pallister Killian syndrome (tetrasomia 12p).
MR imaging enabled the assessment of fetal structures and was helpful in
revealing abnormalities of the uteroplacental unit and the mother. In one
mother, sonography revealed an ectopic pregnancy and findings that were
suggestive of myometrium rupture. This could not be confirmed during
laparoscopy. MR images revealed a right cornual pregnancy with the amniotic
sac enveloped in a thin unruptured layer of myometrium; the uterus held a
large hematoma. Spontaneous abortion followed a few days later. In one woman,
symphisitis was diagnosed on MR imaging, explaining clinical symptoms
suggestive of impending uterine rupture. Septate uteri were detected in two
women (Fig.
6A,6B,6C).
A myomatous uterus (already diagnosed on sonography) was the only pathologic
finding in another patient who was referred for repeated elevated
-fetoprotein levels; follow-up in this patient revealed normal
findings.
|
|
|
In a patient with maternal cervical carcinoma, MR imaging excluded invasion of the parametria and enlarged lymph nodes. Conization of the cervix and subsequent cerclage after Shirodkar was performed during the 12th week of pregnancy. Primary cesarean section and subsequent hysterectomy were performed at term. In another patient with a sonographically detected 12-cm cystic ovarian lesion, MR imaging excluded metastatic spread of disease. At term, a cesarean section was performed and the ovarian lesion was excised, revealing a mucinous tumor. An abscess was excluded in one patient who underwent laparoscopic appendectomy during pregnancy. In these women, prenatal MR imaging and follow-up of the infants after birth revealed normal findings.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Single-shot fast spin-echo MR imaging can be considered a variant of rapid relaxation-enhanced MR imaging [18, 22, 23], in which phase encodings for sections are collected after a single radiofrequency excitation. Additionally, only more than half the K-space is sampled, reducing data acquisition time by one half.
Data acquisition for a single strongly T2-weighted MR image requires less than 1 sec and provides excellent image quality devoid of motion artifacts. Because only one excitation pulse is used to tip the longitudinal magnetization into the transverse plane, the TR is essentially infinite, eliminating all T1-contrast enhancement from the image. Therefore, considerably more T2-weighting is achieved than with conventional or fast spin-echo sequences. Because the magnetization is repeatedly refocused using 180° pulses, single-shot fast spin-echo imaging is less sensitive to susceptibility differences than echoplanar sequences.
In this study, all examinations were of diagnostic quality, free from motion artifacts. Our results indicate that ultrafast T2-weighted single-shot fast spin-echo MR imaging is well suited for detecting maternal disease in pregnant patients and confirms the results of Levine and Edelman [18], Levine et al. [19], and Levine and Barnes [21] that suggest that ultrafast fetal MR imaging provides a detailed and reproducible evaluation of normal fetal anatomy. Furthermore, our study supports the findings of Yamashita et al. [20], who used MR imaging in a smaller group of patients and suggest that fetal ultrafast MR imaging is potentially useful for prenatal assessment.
MR imaging performed best in the evaluation of cerebral abnormalities. Its diagnostic accuracy was superior to that of sonography in some patients. Interpretation of MR images resulted in the correction of standing sonography diagnoses in three patients. Additionally, a previously unknown watershed infarction was detected on MR imaging. These findings confirm the results of a study by Levine et al. [24], in which additional information provided by MR imaging altered counseling in seven (39%) of 18 pregnant women with fetuses affected by central nervous system anomalies. However, MR imaging of the extremities and the heart is currently limited because of missing flow information and the lack of real-time information. These problems might be overcome by future technical advances focusing on interactive rapid imaging.
We do not question the primary role of sonography in prenatal imaging. Lack of a statistically significant difference in lesion detection between sonography and MR imaging reflects the strong selection bias toward severe and cerebral deformities, which were referred for MR imaging.
Sonography revealed deformities (e.g., syndactily or ventricular septal defect) that were less obvious and sometimes overlooked on MR imaging. In the gross malformation syndromes presented in this study, MR imaging played only a subsidiary role; however, in isolation, MR imaging would be crucial in prenatal diagnosis. Also, sonography is less expensive than MR imaging and is a readily available real-time investigation that can provide Doppler blood flow information.
An advantage of MR imaging is that it is not dependent on the prenatal diagnostician's presence. The images can be stored for subsequent analysis or transmitted to a specialist for a second opinion. This advantage could be useful in difficult cases and at isolated or small institutions. Furthermore, MR imaging might acquire an important role in situations (e.g., maternal obesity, oligohydramnios, or fetal head engagement) in which sonographic findings are impaired during late pregnancy.
This study suggests that ultrafast sequences will further increase the clinical indications for in vivo fetal imaging. However, MR imaging should be restricted to situations in which sonographic findings are ambiguous or impaired.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. A. Barth and E. Rubesova Fetal Magnetic Resonance Imaging: Anomalies of the Neck, Chest, and Abdomen NeoReviews, August 1, 2007; 8(8): e313 - e335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Kazan-Tannus, V. Dialani, M. L. Kataoka, G. Chiang, H. A. Feldman, J. S. Brown, and D. Levine MR Volumetry of Brain and CSF in Fetuses Referred for Ventriculomegaly Am. J. Roentgenol., July 1, 2007; 189(1): 145 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O.A. Glenn and J. Barkovich Magnetic resonance imaging of the fetal brain and spine: an increasingly important tool in prenatal diagnosis: part 2. AJNR Am. J. Neuroradiol., October 1, 2006; 27(9): 1807 - 1814. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. A. Glenn, R. B. Goldstein, K. C. Li, S. J. Young, M. E. Norton, R. F. Busse, J. D. Goldberg, and A. J. Barkovich Fetal Magnetic Resonance Imaging in the Evaluation of Fetuses Referred for Sonographically Suspected Abnormalities of the Corpus Callosum J. Ultrasound Med., June 1, 2005; 24(6): 791 - 804. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Farhataziz, J. E. Engels, R. M. Ramus, M. Zaretsky, and D. M. Twickler Fetal MRI of Urine and Meconium by Gestational Age for the Diagnosis of Genitourinary and Gastrointestinal Abnormalities Am. J. Roentgenol., June 1, 2005; 184(6): 1891 - 1897. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hassibi, N. Farhataziz, M. Zaretsky, D. McIntire, and D. M. Twickler Optimization of Fetal Weight Estimates Using MRI: Comparison of Acquisitions Am. J. Roentgenol., August 1, 2004; 183(2): 487 - 492. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Malinger, D. Lev, T. Lerman-Sagie, D. Levine, and P. Barnes Fetal Central Nervous System: MR Imaging versus Dedicated US--Need for Prospective, Blind, Comparative Studies [letter] * Drs Levine and Barnes respond: Radiology, July 1, 2004; 232(1): 306 - 307. [Full Text] [PDF]
|
|
|
|
|
|
F. V. Coakley, O. A. Glenn, A. Qayyum, A. J. Barkovich, R. Goldstein, and R. A. Filly Fetal MRI: A Developing Technique for the Developing Patient Am. J. Roentgenol., January 1, 2004; 182(1): 243 - 252. [Full Text] [PDF]
|
|
|
|
|
|
J. T. Caire, R. M. Ramus, K. P. Magee, B. K. Fullington, D. H. Ewalt, and D. M. Twickler MRI of Fetal Genitourinary Anomalies Am. J. Roentgenol., November 1, 2003; 181(5): 1381 - 1385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Levine, P. D. Barnes, R. R. Robertson, G. Wong, and T. S. Mehta Fast MR Imaging of Fetal Central Nervous System Abnormalities Radiology, October 1, 2003; 229(1): 51 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Kubik-Huch, S. Wildermuth, L. Cettuzzi, A. Rake, B. Seifert, R. Chaoui, and B. Marincek Fetus and Uteroplacental Unit: Fast MR Imaging with Three-dimensional Reconstruction and Volumetry—Feasibility Study Radiology, May 1, 2001; 219(2): 567 - 573. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
