Hannon, Therese BMBS; Tennant, Peter W. G. MSc; Rankin, Judith PhD; Robson, Stephen C. MD
Sonographic evaluation of the fetal cerebral ventricles is recommended as part of the fetal anomaly screening program offered to all women between 18 0/7 and 20 6/7 weeks of gestation in England.1 After 15 weeks of gestation, the ventricular atrium is of a constant size, with a normal width defined as less than 10 mm.2 Atrium measurements above this cutoff are defined as ventriculomegaly, which is subdivided into mild (10–12 mm), moderate (13–15 mm), and severe (greater than 15 mm).3 We recently have reported the prevalence, natural history, and clinical outcome of mild to moderate ventriculomegaly in the north of England.4
Prenatally detected severe ventriculomegaly is known to have a poor prognosis both in terms of survival and neurodevelopmental outcome.5 Prognosis is related to the underlying etiology, which includes obstructive hydrocephalus and cerebral atrophy,6 and the presence of additional anomalies.7 Counseling families when the fetus is affected with severe ventriculomegaly is challenging, and a high proportion of families opt for termination of pregnancy (which is legal in the United Kingdom after 24 weeks of gestation if there is thought to be substantial risk of serious handicap).
Data on the natural history and early outcome of prenatally detected severe ventriculomegaly are still very limited. The aim of this prospective study is to estimate the prevalence, associated anomalies, progression, and clinical outcome of prenatally diagnosed severe ventriculomegaly in a large unselected population.
PATIENTS AND METHODS
The Northern Congenital Abnormality Survey was established in 1985 and collects data on congenital abnormalities arising within the population of the north of England (northeast England and north Cambria). This is a well-defined geographical area with a population of approximately 3 million and an annual live birth rate of 30,000. Data are collected on anomalies arising in pregnancies resulting in late miscarriage (20 weeks of gestation or greater), termination of pregnancy for fetal anomaly after a prenatal diagnosis at any gestation, stillbirth (24 weeks of gestation or greater), and live births.
Details of data collection have been published previously.8 Cases of fetal ventriculomegaly are identified from multiple sources, including regional obstetric departments, the tertiary fetal medicine unit, cytogenetic laboratories, and pediatric and pathology departments. Cross-validation with the genetics and pediatric cardiology databases and with the Northern Perinatal Mortality Survey9 (which records information on all deaths in the first year of life) allows maximal case ascertainment. The Northern Congenital Abnormality Survey adheres to the European Surveillance of Congenital Anomalies inclusion criteria10 (www.eurocat-network.eu/content/EUROCAT-Guide-1.3.pdf) and codes anomalies using the World Health Organization International Classification of Disease, 10th Revision.11 Up to six anomalies can be recorded per case.
Standard measurement of ventricular atrium width, as described by Cardoza et al,2 was included in the scanning protocol for anomaly scans undertaken at all units in north of England from 1994 onward. Standardized guidelines for the management of cases of fetal ventriculomegaly (atrium measurement of 10 mm or greater) were agreed and units asked to report cases to the Northern Congenital Abnormality Survey. All cases of ventriculomegaly detected between October 1, 1994, and December 31, 2008, were identified from the Northern Congenital Abnormality Survey and cross-referenced to cases seen at the regional tertiary fetal medicine unit (Royal Victoria Infirmary, Newcastle Upon Tyne, UK). This was undertaken using the fetal medicine database (PIA Viewpoint) by searching the field of Vp for values of 10 mm or greater.
All singleton pregnancies with confirmed severe ventriculomegaly, defined as a lateral atrium width of 15 mm or greater at any week of gestation, formed the population-based case series. Cases in multiple pregnancies were excluded, because the etiology of ventriculomegaly may relate to the complications of twinning.12
According to the regionally agreed guideline, a specialist in obstetric ultrasonography or a fetal medicine specialist undertook confirmation of fetal ventriculomegaly and assessment for associated anomalies within 7 days of the initial diagnosis. Cases that were unconfirmed, either because a specialist did not perform a detailed scan or because a specialist scan did not confirm ventriculomegaly, were excluded from the analysis.
All women with a confirmed diagnosis were offered fetal karyotyping and infection screening to detect toxoplasmosis and cytomegalovirus. After 2005 (when it became available), fetal magnetic resonance imaging (MRI) was offered to a limited number of women seen in the tertiary fetal medicine center in whom it was thought that MRI may alter their management. Ultrasonographic scans were performed 2–4 weeks after the initial diagnosis and then 4–6 weekly until either resolution (atrium measurement less than 10 mm) or delivery.
Cases were classified into prenatal hierarchical categories according to the presence of: 1. an abnormal karyotype; 2. any European Surveillance of Congenital Anomalies-defined structural anomaly10; 3. cytomegalovirus or toxoplasmosis infection; 4. fetal growth restriction, defined as fetal abdominal circumference less than the third centile for gestation; and 5. no additional prenatal findings.
Cases were classified into the highest matching category (eg, an aneuploid fetus with growth restriction was assigned to category 1 only). Prenatally isolated cases of severe ventriculomegaly were defined as those with no additional complications (ie, category 5 only). All other cases were considered nonisolated.
After delivery, a cranial ultrasonographic examination was recommended between 7 and 10 days of age. Outcome information was obtained from pediatric and radiologic records. In cases in which the pregnancy did not result in a live birth, histopathology and cytogenetic records were reviewed where available.
Continuous variables were summarized into medians and interquartile ranges as a result of divergence from the normal distribution (examined using the Shapiro-Wilk W test). The 95% confidence intervals (CIs) for prevalence rates and proportions were estimated using the Clopper-Pearson (or “exact”) method, which takes values directly from the binomial distribution.13 Demographic and outcome variables were compared between cases of isolated and nonisolated ventriculomegaly using the χ2 test (for categorical variables with five or more expected in each category), Fisher’s exact test (for categorical variables with less than five expected in at least one category), and Mann-Whitney U test (for continuous variables). Changes in prevalences or proportions were examined by the χ2 test for trend.14 The difference in atrium measurement between groups, adjusting for gestational age, was examined by linear regression. The association between atrium measurement and neonatal death, adjusting for isolated status and gestational age, was examined by logistic regression. The probability of spontaneous fetal loss was estimated using the Kaplan-Meier method. Cases were entered at their gestational age at scan and “failed” (for spontaneous fetal deaths) or were censored (for terminations of pregnancy) at their gestational age at delivery. Live births were treated as having survived until the latest gestational age. Change in atrium measurement during pregnancy was modeled using a multilevel approach. Measurements were nested within individuals and modeled by linear regression with random intercepts and continuous gestational age and prenatal isolation included as fixed effects. Parameters were estimated using the Metropolis-Hastings Markov Chain Monte Carlo algorithm,15 assuming diffuse uniform priors, with a burn-in sample of 500 and a monitoring sample of 50,000. Ninety-five percent credible intervals were derived from each parameter's posterior distribution from which P values were also approximated. Statistical analyses were performed using Stata 11.2 and MLwiN 2.25. P<.05 was considered statistically significant.
The Northern Congenital Abnormality Survey, as part of the British Isles Network of Congenital Anomaly Registers, has exemption from the UK National Information Governance Board for Health and Social Care from a requirement for consent for inclusion on the register and has ethics approval (09/H0405/08) to undertake studies involving use of the data.
Figure 1 details the derivation of the study sample. During the study period, 914 cases of ventriculomegaly were identified to the Northern Congenital Abnormality Survey. There were 454,080 registered births in this period, giving a prevalence of suspected ventriculomegaly of 20.1 per 10,000 births (95% CI 18.8–21.5). After applying the study’s inclusion and exclusion criteria, 157 singleton cases of severe ventriculomegaly were identified from 441,247 births, a prevalence of 3.6 per 10,000 singleton births (95% CI 3.0–4.2). There was no change in the prevalence of severe ventriculomegaly over time (the prevalence was 3.4 per 10,000 [95% CI 2.7–4.3] and 3.7 per 10,000 [95% CI 2.9–4.6] during 1994–2001 and 2002–2008, respectively, P for trend=.58).
Derivation and pregn...Image Tools
The maternal and pregnancy characteristics of the study population are summarized in Table 1. Of the 157 cases of severe ventriculomegaly, 79 (50.3%, 95% CI 42.2–58.4) had no additional complications (isolated group), including 85 (54.1%, 95% CI 46.0–62.1) with no associated structural fetal anomalies. Overall, the proportion of cases with additional prenatally identified complications had not significantly changes over the study period (48.1% [95% CI 36.5–59.7] and 52.5% [95% CI 41.0–63.8] during 1994–2001 and 2002–2008, respectively, P for trend=.12). Median maternal age was 27 (interquartile range 23–32, range 17–42) years with a fetal male:female ratio of 1.07:1. Although gestational age at presentation was broadly similar between isolated and nonisolated cases (P=.16 overall), a small, but significantly larger, number of isolated cases entered the cohort at or after 30 weeks of gestation (33% compared with 17%, P=.02). Overall, 60 (38.2%, 95% CI 30.6–46.3) cases presented at or after 24 weeks of gestation. The median atrial measurement at presentation was higher in the isolated (20 [interquartile range 17–24, range 15–46] mm) group compared with the nonisolated group (18 [interquartile range 16–21, range 15–65] mm, P<.001). However, this difference was not statistically significant after adjustment for gestational age at first scan (P=.21). Fifty-seven (36.3%, 95% CI 28.8–44.3) women accepted the offer of prenatal karyotyping. Neither the decision to undergo karyotyping nor the timing of this differed significantly between the isolated and nonisolated ventriculomegaly groups (P=.32).
Details of the 78 nonisolated cases with additional complications are shown in Table 2. Chromosomal anomalies were detected in five cases (3.2%, 95% CI 1.0–7.3). Of these, four were trisomies, and the remaining case was a translocation. Structural anomalies were detected in 67 fetuses (42.7%, 95% CI 34.8–50.8), and of these, 22 (14%, 95% CI 9.0–20.4) had multiple anomalies (ie, additional structural anomalies outside the nervous system). Anomalies of the central nervous system were the most common, identified in 53 cases (33.8%, 95% CI 26.4–41.7); 27 (17.2%, 95% CI 11.6–24.0) had spina bifida, 15 (9.6%, 95% CI 5.4–15.3) had agenesis of the corpus callosum, and 12 (7.6%, 95% CI 4.0–13.0) had other central nervous system anomalies. Ten cases had cardiovascular anomalies (6.4%, 95% CI 3.1–11.4) and seven had orofacial clefts (4.5%, 95% CI 1.8–9.0). Four cases (2.5%, 95% CI 0.7–6.4) were associated with fetal growth restriction. Maternal and fetal testing identified congenital toxoplasmosis in two cases (1.3%, 95% CI 0.2–4.5).
Serial prenatal atrium measurements were available for 53 fetuses. Thirteen cases (24.5%, 95% CI 13.8–38.3) had a lower follow-up measurement, including nine (17.0%, 95% CI 8.1–29.8) in which the classification had changed; five (9.4%, 95% CI 3.1–20.7) regressed to moderate ventriculomegaly, three (5.7%, 95% CI 1.2–15.7) to mild ventriculomegaly, and one (1.9%, 95% CI less than 0.1–10.1) to a normal atrium size. Of the 44 cases that remained severe throughout, the median change between the first and last scan was 4.1 (interquartile range 1.0–9.0, range −4.0 to 28.6) mm over a median time period of 3.5 (interquartile range: 1.5–9.5, range 0.0–17.0) weeks.
Change in atrium measurement for all 53 cases with follow-up data was estimated by multilevel linear regression. On average, the atrium measurement increased by 0.67 mm per week (95% credible interval 0.48–0.86). The rate of increase was borderline significantly greater in isolated cases (0.76 [95% credible interval 0.54–0.98] mm per week) compared with nonisolated cases (0.40 [95% credible interval 0.16–0.87] mm per week, P=.051). In the 22 cases that received their first scan before 27 weeks of gestation, the median change between the first and last scans was 6.5 (interquartile range 2.0–13.0, range −4.0 to 28.6) mm over a median time period of 8.5 (interquartile range 1.0–14.0, range 1.0–17.0) weeks. On average the atrium measurement increased by 0.92 mm per week (95% credible interval 0.68–1.16), but there was no apparent difference in the rate of increase between isolated and nonisolated cases (0.96 [95% credible interval −3.01 to 4.96] mm difference, P=.58).
Thirty-seven cases with a follow-up atrium measurement were live-born, among whom an association was observed between final atrium measurement and mode of delivery (P=.015) with the odds of an elective cesarean delivery being increased by 11% for each additional millimeter in atrial width (odds ratio [OR] 1.11, 95% CI 1.02–1.21). Change in atrium measurement, however, did not predict mode of delivery, irrespective of adjustment for final atrium measurement or whether or not the case was isolated (adjusted P=.35).
The pregnancy outcomes of the isolated and nonisolated groups are shown in Table 3. Sixty of the 78 women (76.9%, 95% CI 66.0–85.7) with nonisolated fetal severe ventriculomegaly opted to terminate their pregnancy compared with 41 of 79 women (51.9%, 95% CI 40.4–63.3) with apparently isolated ventriculomegaly (P=.001). The cumulative probability of spontaneous fetal death (as estimated by Kaplan-Meier, to account for censorship of cases ending in elective terminations) was estimated to be 15.3% (95% CI 3.9–50.2) in the nonisolated group and 11.1% (95% CI 1.6–56.7) in the isolated group (P=.35). There was a total of 53 live births (33.8%, 95% CI 26.4–41.7), 37 in the prenatally isolated group (46.8%, 95% CI 35.5–58.4) and 16 (20.5%, 95% CI 12.2–31.2) in the nonisolated group (P<.001).
Additional postnatal anomalies were identified in 22 of the 79 cases that appeared to be isolated prenatally (27.8%, 95% CI 18.3–39.1), including 12 (29.3%, 95% CI 16.1–45.5) live-born neonates (Table 4). In addition, one case of congenital cytomegalovirus infection was identified. Baseline atrium measurement did not predict the presence of undetected anomalies in the 79 cases with prenatally isolated severe ventriculomegaly (OR 1.03 [95% CI 0.95–1.11] per mm, P=.47). Eleven of 53 live-born neonates died (20.8%, 95% CI 10.8–34.1), five of 16 neonates in the group with prenatally detected associated anomalies (37.5%, 95% CI 11.0–58.7) and six of 37 neonates with isolated severe ventriculomegaly (16.2%, 95% CI 6.2–32.0) (P=.22). In only one case was this after an active decision to provide palliative care only after delivery at 27 weeks of gestation. Despite small numbers, the risk of neonate death appeared to increase with increasing baseline atrium measurement, although the association was below the nominal significance level (crude OR 1.05 [95% CI 0.98–1.13; P=.14, adjusted for isolated status and gestational age OR 1.07 [95% CI 0.99–1.15] per mm, P=.09). Among the 42 live-born cases with progression data, change between first and last atrium measurement did not predict neonatal death, whether before or after adjustment for isolated status and gestational age at delivery (adjusted P=.90). Although this study is the largest of its type, it remains small in absolute terms. There were very few cases of certain outcomes or in certain subgroups. Several estimates thus carry a high degree of uncertainty and nonsignificant P values should not be viewed as evidence of no effect. To reduce the risk of false-negative conclusions, we reported the point estimates and CIs for all nonsignificant associations with P<.1.
This population-based study collected data on prenatally diagnosed fetal ventriculomegaly from all obstetric units in the north of England. Most cases were detected during routine ultrasonographic screening with a median gestational age at confirmation of 21 weeks, much earlier than reported in previous series.5,16 The data collection methods used by the Northern Congenital Abnormality Survey meant that complete data on pregnancy and postnatal outcome were available for all cases. This study represents a large series of cases of severe ventriculomegaly, and the ultrasonographic follow-up strategy provides unique information on progression of ventriculomegaly.
The prevalence of severe ventriculomegaly in this population is 3.6 per 10,000 singleton births. This compares with 7.9 per 10,000 cases of mild to moderate ventriculomegaly in the same population.4 Comparisons to prevalence rates reported for other populations is difficult as a result of lack of accurate population-based data. The European Surveillance of Congenital Anomalies reported a prevalence of congenital hydrocephalus of 4.92 per 10,000 births, but prevalence varied considerably among the 17 countries included in this study, possibly reflecting differences in screening policies and definitions of hydrocephalus.17 Because fetal anomaly screening is universally offered and widely accepted by pregnant women in our region, and definitions of ventriculomegaly are consistent, our prevalence data are likely to be an accurate estimate for this stable, predominantly white, northern European population.
The association of severe ventriculomegaly with other anomalies is critical when counseling parents about prognosis. Prior studies have shown that associated anomalies confer increased risks of both neonatal mortality and adverse neurodevelopmental outcome.7,16,18 The overall rate of associated anomalies in fetuses with severe ventriculomegaly (49.7%) was similar to that previously reported in fetuses with mild to moderate ventriculomegaly (44.8%, P=.31).4 The rate of prenatally diagnosed chromosomal anomalies in the present series (3.2%) is somewhat lower than reported previously (5–8.3%).5,16 Because not all cases were karyotyped, this represents a minimum estimate of aneuploidy rate. This compares with a chromosomal anomaly prevalence of 14.1% among those with mild-to-moderate ventriculomegaly,4 confirming the previously reported preponderance of chromosomal anomalies among mild to moderate cases.7,19 Structural anomalies were detected in 42.7% of cases by prenatal imaging. In agreement with the series by Gaglioti et al,18 these were most commonly anomalies of the nervous system.
Consistent with previous reports,16,20 we found that severe ventriculomegaly was associated with a high risk of neonatal mortality (16.2%), even in the absence of prenatally detected associated anomalies. This is significantly higher (P=.003) than the 3.0% neonatal mortality rate we reported for isolated mild to moderate ventriculomegaly.4 Perhaps surprisingly, despite specialist ultrasonographic examination and, lately, fetal MRI, we report a significant proportion (27.8%) of cases thought to have isolated ventriculomegaly who had associated anomalies detected after birth, predominantly in the fetal brain. This figure is in broad agreement with other authors,5,21 and there was no evidence of a change in this rate over the period of the study, even after the introduction of fetal MRI. Advanced ultrasound techniques22,23 and the emergence of MRI as a technique to assess the fetal brain have been reported to improve detection of associated anomalies.24–26 Importantly, recent evidence suggests that MRI provides accurate diagnostic information on brain anomalies at 20–24 weeks of gestation and that there is no diagnostic advantage to either delaying imaging once ventriculomegaly is detected at routine screening or reimaging in the third trimester.27 It is possible that as experience with this technique is gained and its use becomes routine in clinical practice, the detection rate of associated anomalies may increase over time. Prenatal knowledge of associated infection, karyotype anomalies, and structural anomalies is essential and allows parents to make informed choices about continuation or interruption of their pregnancy. Significantly more parents with “nonisolated” severe ventriculomegaly opted to interrupt the pregnancy in comparison to parents in the “apparently isolated” group, emphasizing the importance of accurate prenatal diagnosis.
The current study provides unique information on the natural history of severe ventriculomegaly in utero. In almost one in five cases, atrium size decreased to a lesser classification of ventriculomegaly after an initial confirmed diagnosis. Indeed, one case (with an initial atrium width of 15 mm) completely regressed to normal. The likelihood of delivery by elective cesarean delivery was dependent on the final atrium size. Importantly, although there was a potential association between risk of neonatal death and increasing baseline atrium size, the rate of change in atrium measurement did not predict neonatal death in either the isolated or nonisolated group.
In conclusion, this large, population-based case series provides unique data that can be used to counsel parents that there is a significant risk of postnatal additional anomaly detection and of neonatal death even in apparently isolated ventriculomegaly. These outcomes are not predicted by an increase in atrial size antenatally.
The strengths of the study include the high case ascertainment and the complete pregnancy outcome data. Although this represents a large series, the number of live births is still relatively small. As yet there are no data on the neurodevelopmental outcome of the neonates who were live-born and survived infancy, and this work is ongoing.
2. Cardoza JD, Goldstein RB, Filly RA. Exclusion of fetal ventriculomegaly with a single measurement: the width of the lateral ventricular atrium. Radiology 1988;169:711–4.
3. Wax JR, Bookman L, Cartin A, Pinette MG, Blackstone J. Mild fetal cerebral ventriculomegaly: diagnosis, clinical associations, and outcomes. Obstet Gynecol Surv 2003;58:407–14.
4. Sethna F, Tennant PWG, Rankin J, C Robson S. Prevalence, natural history, and clinical outcome of mild to moderate ventriculomegaly. Obstet Gynecol 2011;117:867–76.
5. Kennelly MM, Cooley SM, McParland PJ. Natural history of apparently isolated severe fetal ventriculomegaly: perinatal survival and neurodevelopmental outcome. Prenat Diagn 2009;29:1135–40.
6. D'Addario V, Pinto V, Di Cagno L, Pintucci A. Sonographic diagnosis of fetal cerebral ventriculomegaly: an update. J Matern Fetal Neonatal Med 2007;20:7–14.
7. Nicolaides KH, Berry S, Snijders RJ, Thorpe-Beeston JG, Gosden C. Fetal lateral cerebral ventriculomegaly: associated malformations and chromosomal defects. Fetal Diagn Ther 1990;5:5–14.
8. Richmond S, Atkins J. A population-based study of the prenatal diagnosis of congenital malformation over 16 years. BJOG 2005;112:1349–57.
10. European Surveillance of Congenital Anomalies. EUROCAT Guide 1.3 and reference documents. Instructions for the registration and surveillance of congenital anomalies. Available at: http://www.eurocat-network.eu/content/EUROCAT-Guide-1.3.pdf
. Retrieved June 19, 2012. Issued September 2005. revised May 2012.
11. World Health Organization. International statistical classification of diseases and related health problems. 10th revision. Vol 2; 2010. Available at: http://www.who.int/classifications/icd/en/
. Retrieved June 19, 2012.
12. Tarui T, Khwaja OS, Estroff JA, Robinson JN, Gregas MC, Grant PE. Altered fetal cerebral and cerebellar development in twin-twin transfusion syndrome. AJNR Am J Neuroradiol 2012;33:1121–6.
13. Clopper C, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934;26:404–13.
14. Armitage P. Tests for linear trends in proportions and frequencies. Biometrics 1955;11:375–86.
16. Breeze ACG, Alexander PMA, Murdoch EM, Missfelder-Lobos HH, Hackett GA, Lees CC. Obstetric and neonatal outcomes in severe fetal ventriculomegaly. Prenat Diagn 2007;27:124–9.
17. Garne E, Loane M, Dolk H, De Vigan C, Scarano G, Tucker D, et al.. Prenatal diagnosis of severe structural congenital malformations in Europe. Ultrasound Obstet Gynecol 2005;25:6–11.
18. Gaglioti P, Danelon D, Bontempo S, Mombrò M, Cardaropoli S, Todros T. Fetal cerebral ventriculomegaly: outcome in 176 cases. Ultrasound Obstet Gynecol 2005;25:372–7.
20. Graham E, Duhl A, Ural S, Allen M, Blakemore K, Witter F. The degree of antenatal ventriculomegaly is related to pediatric neurological morbidity. J Matern Fetal Med 2001;10:258–63.
21. Lee CS, Hong SH, Wang KC, Kim SK, Park JS, Jun JK. Fetal ventriculomegaly: prognosis in cases in which prenatal neurosurgical consultation was sought. J Neurosurg 2006;105(suppl):265–70.
22. Cohen-Sacher B, Lerman-Sagie T, Lev D, Malinger G. Sonographic developmental milestones of the fetal cerebral cortex: a longitudinal study. Ultrasound Obstet Gynecol 2006;27:494–502.
23. Malinger G, Kidron D, Schreiber L, Ben-Sira L, Hoffmann C, Lev D, et al.. Prenatal diagnosis of malformations of cortical development by dedicated neurosonography. Ultrasound Obstet Gynecol 2007;29:178–91.
24. Griffiths PD, Reeves MJ, Morris JE, Mason G, Russell SA, Paley MN, et al.. A prospective study of fetuses with isolated ventriculomegaly investigated by antenatal sonography and in utero MR imaging. AJNR Am J Neuroradiol 2010;31:106–11.
25. Morris JE, Rickard S, Paley MNJ, Griffiths PD, Rigby A, Whitby EH. The value of in-utero magnetic resonance imaging in ultrasound diagnosed foetal isolated cerebral ventriculomegaly. Clin Radiol 2007;62:140–4.
26. Rickard S, Morris J, Paley M, Griffiths P, Whitby E. In utero magnetic resonance of non-isolated ventriculomegaly: does ventricular size or morphology reflect pathology? Clin Radiol 2006;61:844–53.
27. Griffiths PD, Morris JE, Mason G, Russell SA, Paley MN, Whitby EH, et al.. Fetuses with ventriculomegaly diagnosed in the second trimester of pregnancy by in utero MR imaging: what happens in the third trimester? AJNR Am J Neuroradiol 2011;32:474–80.
Figure. No available...Image Tools