| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
1 From the Department of Radiology, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215. Received and accepted January 31, 2001. Address correspondence to the author (e-mail: dlevine@caregroup.harvard.edu).
Index terms: Editorials • Fetus, MR, 856.121416 • Magnetic resonance (MR), three-dimensional, 856.12141 • Pregnancy, MR, 856.121416
Magnetic resonance (MR) imaging has a number of contributions to make in the evaluation of the at-risk pregnancy. In addition to the superb depiction of fetal anatomy and anomalies with MR imaging, the data obtained with fast MR imaging techniques can be used to assess the volume of the fetus and supporting structures. The expectation is that fetal weight estimates based on fetal volume determinations will be more accurate than those obtained with ultrasonography (US). US estimates of fetal weight are reasonably accurate for the majority of the fetal population; however, this accuracy declines at the extremes of weight, in cases of intrauterine growth restriction or macrosomia, when precision in US biometry is most critical.
Both fetal macrosomia and intrauterine growth restriction increase the risks of perinatal morbidity and mortality and of long-term neurologic and developmental disorders (1,2). Identification of intrauterine growth restriction after 37 weeks of gestation is an indication for delivery to reduce the chance of fetal mortality (3). Similarly, the diagnosis of macrosomia frequently leads to delivery by means of cesarean section to decrease the risk of failed vaginal delivery due to cephalopelvic disproportion and the risk of shoulder dystocia. Sonographic estimation of fetal weight is performed with a variety of equations by using combinations of some or all of the following measurements: fetal biparietal diameter, head circumference, abdominal circumference, and femur length (4,5). In general, the mean absolute error of US-predicted birth weight varies from 6% to 12% of the actual birth weight, and 40%–75% of estimates are within 10% of actual birth weight (5–8). Compared with birth weight, US estimates overestimate low birth weight and underestimate high birth weight (9,10). MR imaging has the promise of being less affected by patient body habitus (unless the patient cannot fit into the magnet bore); instead of biometric measurements being used for estimation of weight, a true fetal mass can be assessed with volumetric acquisition that shows the entire fetal contour.
A number of methods have been described for obtaining the data with which fetal volumetry can be performed (Table) (11–16). In the current issue of Radiology, Dr Kubik-Huch and colleagues (15) describe two-dimensional MR imaging performed with a half-Fourier single-shot fast spin-echo technique in which fetal, amniotic fluid, and placental volumes were manually segmented and reconstructed with automatic volume calculations. They validated the method in vitro by means of fluid displacement and direct weight of a fetal specimen. They then studied 22 fetuses at 16–41 weeks. Weight estimates with US and MR imaging were compared. In addition, findings in six cases in which MR imaging was performed within 1 week of delivery were compared with birth weight.
|
As Kubik-Huch et al (15) state, the results of their study demonstrate the feasibility of fetal, amniotic fluid, and placental volumetry on the basis of fast MR imaging data sets. The demonstration of feasibility is a good first step and adds to previous reports that document differing techniques for fetal weight estimation with MR imaging. The next questions to address are whether the derived data are clinically important and whether the data can be efficiently obtained in a busy obstetric imaging practice.
Thus, there are three major challenges regarding fetal volumetry that need to be addressed before this technique can be used in clinical practice. The first of these challenges is dependent on advances in MR imaging hardware and software to allow imaging protocols fast enough for a fetal volume to be obtained on the order of seconds. The relative ease of covering the smaller fetal volume in the second trimester (compared with the larger fetal mass in the third trimester) is offset by the large body movements that can occur in the second trimester, which make volume acquisition and reconstruction difficult. Similarly problematic is the relatively larger fetal volume in the third trimester, which requires more time to cover and thus allows fetal motion to degrade images. Sequence acquisition time is crucial in these endeavors because longer acquisitions are more likely to be degraded by fetal motion, which leads to inaccuracies in volume determination.
The non–echo-planar imaging protocols described to date for fetal volumetry (Table) all require at least 1 minute of imaging, which allows fetal motion to occur between section acquisitions. In our laboratory, we have performed three-dimensional fetal volume acquisitions with a true fast imaging with steady-state precession, or true FISP, technique during a maternal breath hold in less than 30 seconds (17), but we have not yet applied the volume data to estimations of fetal weight. More work is needed in this area to derive MR imaging pulse sequences that allow the volume of the fetus to be rapidly and reproducibly assessed.
The second challenge for fetal volumetry is determination of the appropriate method for estimating fetal volume. This can be based on manually tracing fetal outlines, which is very time-consuming, as evidenced by the 6 hours per case in the study of Kubik-Huch et al (15). The use of software to decrease the human time needed to select fetal volumes allows fetal segmentation in as little as 10 minutes (11). With the Cavalieri method of stereologic examination combined with point counting, wherein the volume is estimated on the basis of the sum of the transected areas of the sections through the structure multiplied by the gap between the MR imaging sections, case analysis can occur in as little as 5 minutes per case (12,13,18). Since the uterine wall, fetal soft tissues, and placenta often have similar signal intensities, completely automated segmentation is not practical currently. Fetal motion may also interfere with automated segmentation since sequential images may have discontinuous fetal borders. The feasibility of MR imaging in daily practice will be greatly influenced by the ease with which segmentation and three-dimensional modeling can be accomplished.
The third challenge for fetal volumetry is definition of the relationship between fetal volume and weight. It is not possible to extrapolate directly from fetal weight to fetal volume because fetal density varies with factors such as fetal fat, water, muscle, and bone content. Results determined by assuming a mean fetal tissue density coefficient of 1.0 g/cm3 underestimate fetal weight at term (11,12, 15,16). Formulas extracted from small populations of fetuses at term derive density coefficient values that vary from 1.03% to 1.07% (11,16). The equations derived from fetuses at term may be of no use in the second trimester, however, since the proportion of fetal tissues is different. This is of particular importance when a formula such as that derived by Baker et al (16) is applied to fetuses remote from term. This formula has a correction factor of 0.12 kg that is unlikely to be applicable in the second trimester, since such a constant has a greater percentage influence on weight calculated in early pregnancy.
Larger study populations in the second and third trimesters are needed to determine the relationship between fetal volume and fetal weight with respect to gestational age. In addition to variations in fetal density with gestational age, different populations of fetuses may need to be evaluated differently. Water content decreases with increasing gestation, and fetuses with intrauterine growth restriction have lower water and fat content than do those with appropriate growth. A macrosomic fetus with increased fat will likely have a different weight per unit volume of tissue in comparison with a growth-restricted fetus with little subcutaneous fat. Therefore, variation in fetal density will likely be greatest at extremes of growth abnormalities. US estimates of fetal weight have been based on studies of large populations. Many iterations of fetal weight calculations have been proposed, and many variations of these equations are currently in use. With no universally accepted standard established, there is no doubt that controversy regarding any new method for fetal weight estimation is expected.
Despite these challenges, MR imaging shows great promise for improved diagnosis of macrosomia and intrauterine growth restriction. This is due to the additional information that can be derived from these studies. Studies of MR imaging measurements of liver have shown a single fetal liver volume measurement, performed several weeks before delivery, can help distinguish fetuses subsequently diagnosed as being growth restricted (19). MR imaging techniques are available to assess the fat content in the liver, and these may provide an additional means of fetal assessment.
Oligohydramnios is another finding in intrauterine growth restriction. The typical method to assess amniotic fluid volume is the amniotic fluid index; however, variations in technique hamper the ability to standardize assessment of oligohydramnios. It is possible that MR imaging will allow a more reliable means of assessing amniotic fluid volume.
Information regarding fetal fat (20), functional evaluation of the placenta (21,22), and placental volume assessments will likely be used in combination with other data to better distinguish between the constitutionally small but appropriately grown fetus and the fetus at risk owing to placental insufficiency. Evaluation of the macrosomic fetus may be improved by pursuing MR pelvimetry (23) and by determining fetal shoulder width (23) and fetal fat (20).
Dr Kubik-Huch and colleagues (15) are to be commended for this early work in fetal weight estimation. Challenges for future MR fetal weight determinations include the development of faster imaging techniques to reduce the likelihood of fetal motion during acquisition, three-dimensional MR imaging strategies that allow true volumetric acquisitions, and better software for automated fetal segmentation, and the performance of larger studies to assess the relationship between fetal volume and weight and the way equations that relate volume and weight can be applied to differing fetal populations. The relative merits of MR imaging and US for the assessment of fetuses at the extremes of growth abnormalities remain to be determined in clinical practice.
FOOTNOTES
See also the article by Kubik-Huch et al (pp 567–573 ) in this issue.
REFERENCES
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
