Heron Werner, MD & Pedro Daltro, MD
Clínica de Diagnóstico por Imagem (CDPI) & Instituto Fernandes Figueira (IFF) – FIOCRUZRio de Janeiro – Brazil

Dorothy I. Bulas M.D.
Professor of Radiology and Pediatrics
Children"s National Medical Center
George Washington University Medical Center
111 Michigan Ave, NW,   Washington D.C. 20010

The first study using MRI in gestation was done by Smith in 1983. Later, its use grew progressively in obstetrics and fetal medicine, especially in the evaluation of the central nervous system (CNS) of the fetus. This is due to fact that the fetal CNS is not easily evaluated by ultrasound because of the ossification of the calvaria in the third trimester, besides frequent inappropriate fetal position (Smith et al., 1983; Benson et al., 1991).

Although ultrasound continues to be the most used modality of prenatal exam routine because of its low cost, more equipment availability, safety, high sensibility and good real time capacity of analysis, MRI is potentially good at the morphological evaluation of those fetuses that are not easily studied by ultrasound (Williamson et al. 1989). Vadeyar et al. 2000 registered fetal heart beats before and during MRI exam in 10 fetuses at gestational ages between 37 to 41 weeks, demonstrating no significant alterations.

Safety committees have devised guidelines which say that pregnant employees of MRI centers can go into the exam room without any risk. Nevertheless, they must avoid the exam room during the use of a radiofrequency fields. Some centers still do not recommend these pregnant employees to stay in the exam room during their first trimester of pregnancy. In the USA, they recommend their pregnant employees to be well informed about MRI. Then, they will be able to take their own decision about staying or not in the unit (Brunelle et al., 1998).

Despite the fact that ultrasound has been reaching great progress with the improvement of probes and more recently, with 3D images, such a technology can have its image quality degraded, especially during the third trimester of gestation. The main inherent problem of ultrasound is, according to Poutamo et al. 1999:

• Acoustic skull cap shadow, which prejudices mainly the study of the posterior fossa.
• Presence of the cephalic pole in the maternal pelvis, especially in the transverse insinuations, prejudicing the sagittal evaluation of the cephalic pole. In some of these cases, the problem can be solved by the transvaginal ultrasound probe.
• Distance enlargement between the ultrasound probe and the cerebral structures in the cases of an important hydrocephaly.
• Intracranial hyperechoic images, such as tumors and hemorrhage which can prejudice the rest of the cerebral anatomy evaluation.
• Acoustic shadow presence from the jaw and the base of the cranium which hinders a good evaluation of the cervical region.
• Slight difference of echogenicity between tissues such as the esophagus and the trachea.
• Maternal obesity.
• Strong polyhydramnios or oligohydramnios.

Poutamo et al. 1999 studied 27 fetuses at gestational ages over 20 weeks by ultrasound and MRI. MRI added additional information in 9 (47%) out of 19 cases with CNS anomalies and in 6 (75%) out of 8 cases of cervical anomalies. In 8 cases, the posterior fossa anatomy could not be well evaluated by ultrasound. In the 8 suspected cases of cervical anomaly, the polyhydramnios and the extreme deflexion of the fetal cephalic pole were the main limitation to a good echographic study. In all 8 cases, detailed images at the pharynx and esophagus were not satisfactory by ultrasound. So, MRI provided a better visualization of the posterior fossa, eliminating possible artifacts caused by acoustic shadows, facilitating the diagnosis of the cerebellar hypoplasia and hemorrhagic lesions alongside a good contrast between the trachea, esophagus and soft tissues of the cervical region.

Most recommendations for fetal MRI are related to CNS pathologies. The evaluation of the various stages of fetal cerebral formation proved to be poor by ultrasound. At present, the use of fetal MRI enables us to study the fetal CNS with good sensibility at the end of the second and third trimesters.

The embryogenesis of the brain is already substantially completed at about 20 weeks of gestation. In the first weeks of fetal development, the cerebral hemispheres surface is smooth, because the gyration depends on the conclusion of the neuronal migration, which will be completed near the term of pregnancy. The MRI study of sulcation requires knowledge of the patterns of maturation of sulci and gyri according to fetal age. The MRI results can be compared to the anatomy atlases (Figure 1).

Nevertheless, it is important to know that there are variations among the fetuses in sulcal-gyral development (Garel et al. 1998). According to Chi et al. 1997, the sulci and gyri that should be present at a given gestational age are found in only 25-50% of the brains studied. In the twins, the sulcal maturation has a delay of 2-3 weeks. At the end of the first trimester of pregnancy, primary sulci are present. The gyri appear gradually, being Sylvian fissure the first to appear. Ultimately, calcariae, parietooccipital, cingulate gyri appear (20-22 weeks); the interparietal and temporal (25 weeks), pre-central, post-central and front superior (24-28 weeks) (Figure 2-14).

The secondary and tertiary gyri appear in the last two months of pregnancy (Table 1). The cerebral ventricles can be well defined by MRI. They have a high signal on T2-weighting images. They show morphological and size variations during the pregnancy. There is a physiological relative ventricular enlargement until 25 weeks of gestation which persists at the level of occipital horns until the 30th week. On the ultrasound, the enlargement of the transverse atrial diameter above 10 mm after 25 weeks is considered to be pathological. The cerebral parenchyma can be well studied. The germinal matrix appears as a low signal lying along the lateral ventricular wall on T2-weighted images. The hypointense layers on T2-weighting correspond to high cellularity. The parenchyma appears homogeneous with hyperintensity (T2-W images) after the 28th week. The thalamus appears as hypointense in T2-weighted images by 27 weeks. The all components of the corpus callosum can be seen as a hypo signal on T2-weighted images after 21 weeks of gestation (Figure 15-42).

Although the main indication of fetal MRI are related to CNS pathologies, MRI has shown a large utility in the detection of other fetal anomalies, such as diaphragmatic hernia, urinary and abdominal wall defects (Werner et al., 2001).
The naso-oropharynx and trachea present as a bright structure on T2-weighted because they are filled with amniotic fluid. The thyroid gland is difficult to be identified on T2-weighted, but can be detected on T1-weighted as a moderate hyperintense structure compared with surrounding tissues (Figure 43). The thymus shows intermediate signal intensity in anterior mediastinal area.

The fetal lungs are well demonstrated by MRI thanks to the presence of water in their constitution. Maturation of the lung commences around 24 week's gestation. As the lung develops there is an increase in water content and a rise in the phospholipid concentrations that relate to surfactant production. Therefore, the method allows the evaluation of pulmonary maturity after the 26th week of gestation because the MRI signal intensity could vary in both proteins and lipids concentrations (Fenton 2000) (Figure 44-54).

The volume of the fetal organs can be obtained by MRI. Duncan et al. 1999 studied the pulmonary volume by MRI with the intent to evaluate pulmonary hypoplasia. They found out a direct relation between the pulmonary volume augmentation and the fetal volume augmentation. Such an observation was also found by Lee et al. 1996, who made a pulmonary volume evaluation by 3D ultrasound. Nevertheless, it is important to highlight that despite similarities of results, the ultrasound has its limits in oligohydramnios situations. The ultrasonographic evaluation of the pulmonary hypoplasia is limited by a poor echogenicity differentiation between the pulmonary tissue and the adjacent structures. The MRI pulmonary image is very clear because the pulmonary tissue has a hypersignal in T2-weighting caused by the presence of the amniotic fluid in its interior and not air. Kawashina et al. 2001 studied the fetal pulmonary hypoplasia with regard of the low intensity of the lungs. They studied 23 fetuses with different kinds of anomalies between 18 and 40 weeks of gestation, such as hydrocephaly, urinary malformations, diaphragmatic hernia, skeletal defects, hydrops and growth restriction. The confirmation of pulmonary hypoplasia was made by the postnatal clinic, surgery or pathological study. All fetuses with normal pulmonary development presented a hypersignal in T2 (HASTE). The low intensity signals after the 26th week of gestation were related to pulmonary hypoplasia.
The fetal heart can also be viewed by MRI. A central low intensity area on T2-weighted represents the fetal heart, which can be seen from early in the second trimester. Nevertheless, the architecture of the heart is not well identified given its movement. An adult's heart, if attached to an electrocardiograph, can have an excellent contrast in its structure. Perhaps, we will soon be able to do the same with a fetus. The major vessels of the fetus are also seen as low intensity areas on T2-weighted sequences. This low intensity signals are related to the effect of blood flow on the MRI signal and is depended on the rate of flow, turbulence and the image selected (Figure 44-54).