We thank Professor Timor-Tritsch for his interest and thoughts on our recent Editorial on magnetic resonance imaging (MRI) in fetal medicine, as well as for his ongoing efforts to improve the performance of ultrasound amongst fetal medicine specialists1. We were surprised to learn that he took offence at our article. He interpreted it as if we were claiming that fetal MRI is superior to ultrasound, an argument that he has made before2. However our Editorial did not reach that conclusion. To dispel any confusion about the message we wanted to put across, we would like to summarize here our key points, which we feel are supported by the available data as well as by studies soon to be published.

  • (1)
    By no means do we encourage the mis- or over-use of fetal MRI. This would needlessly drain resources from today's already shrinking healthcare budget, which would be better invested in widely available and high-quality screening programs.
  • (2)
    Fetal MRI is not, and probably will never be, a screening tool, but should be offered when ultrasound is inconclusive or (occasionally) falls short. In our Editorial we gave some examples of situations in which physics limits the performance of ultrasound (oligohydramnios, obesity, malpresentation, bony attenuation…).
  • (3)
    Fetal MRI is practiced by radiologists who, as for any other radiological examination, act according to specific questions posed by clinicians. In other words, the role of fetal MRI will be defined by the caregiver of the pregnant mother, i.e. the obstetrician, or a subspecialist in fetal medicine. Therefore, any use of fetal MRI is by definition controlled by the fetal medicine specialist.
  • (4)
    The present discussion on the two imaging modalities should not be one of antagonism but one that considers MRI as supplementing ultrasound for precise indications. In an effort to solve diagnostic or therapeutic dilemmas, both specialities must liaise and interact and this was precisely the goal of the session at the 2005 ISUOG meeting.
  • (5)
    For many investigational indications, both professions are exploring the circumstances under which one examination may aid in diagnosis, or may not contribute at all. Specialized centers where the two professions already interact intensively are probably the ones best suited to exploring the potential and limitations. Our Editorial did not, at any point, call for radiologists throughout the world to start performing fetal MRI, nor did it plead for routine or unrestricted practice of fetal MRI. Just as not all obstetricians practice fetal ultrasound at the same advanced level, the same will apply to radiologists.
  • (6)
    The quality of fetal MRI is largely operator-dependent, but this applies also to (fetal) ultrasound. It is certainly possible that Professor Timor-Tritsch has seen, more often than he would wish, fetal MR images of substandard quality. In his introduction he confirms that the same goes for sonographers ‘using wrong transducers, wrong scanning routes and incorrect settings’, and this series of unfortunate wrong-doings might be expanded to include ‘wrong timing and wrong indications’. The only solution to avoid these mishaps is training (to overcome the above shortcomings) and research (to explore the potential and limitations of both imaging techniques). Again, this is what ISUOG has been doing, not in the least assisted in this task by people like Professor Timor-Tritsch.

Professor Timor-Tritsch delves deeper into the application of fetal MRI for central nervous system evaluation, an area in which he has undisputed expertise3. We would like to take one of his important points with which, on closer reading, it is clear that we actually agree: the role of fetal MRI early in pregnancy is limited, and indeed with the exception of some research protocols, can be considered non-existent. In reality most fetal MRI examinations are planned based on abnormalities suspected on screening ultrasound examination. In most countries structural abnormality screening is performed between 20 and 24 weeks, a time which is close to the pivotal point of viability, and which poses a tremendous burden on the operators doing any imaging study.

However, we take issue with Professor Timor-Tritsch's suggestion that we have not carefully read the references we cite. We did not claim that fetal MRI was particularly helpful early in pregnancy. In the study by Whitby et al., the diagnostic accuracy of ultrasound and fetal MRI is compared in 100 fetuses with suspected brain anomalies4. In this particular study half of the subjects were assessed prior to 24 weeks, and in 29 cases fetal MRI findings changed the management defined by ultrasound. Among these 29, 12 were assessed between 16 and 24 weeks. It is obviously possible that the numbers and conclusion may have been different if this study had been carried out elsewhere. This, and several other studies on MRI of the fetal brain were performed in centers where expertise in obstetric ultrasound is available4–7. There is certainly no indisputable evidence of the superiority of one imaging technique over the other in these or other studies, but specially designed studies need to be performed to prove such statements.

Furthermore, Professor Timor-Tritsch acknowledges that the undisputed added value of fetal MRI is that it offers information on brain maturation (in the widest sense) which can be lacking on ultrasound. Gyral formation, parenchymal morphology and functional examinations are examples of this, but this information is, by definition, only available late in gestation.

There are some situations in which fetal MRI plays a pivotal role, including when fetuses are due to undergo fetal surgery or ‘ex-utero intrapartum treatment’. Some fetal surgery protocols foresee a clinical role for MRI in conditions such as congenital diaphragmatic hernia, congenital cystic adenomatoid malformation, twin-to-twin transfusion syndrome, upper airway obstructions and sacrococcygeal teratoma8–12. However, given the rarity of these conditions, conclusive evidence of the added value of MRI is lacking.

To conclude, we felt it appropriate to leave the initial question that we raised un- or incompletely answered and were expecting the readership of Ultrasound in Obstetrics and Gynecology to judge for themselves where fetal MRI services stand in the environment in which they work. We use fetal MRI in our departments as a supplement to advanced ultrasound for selected pathology. Within research protocols, imaging modalities are sometimes compared and turn out to have different limitations13. Our Editorial aimed to invite other teams to participate in this ongoing dialogue (rather than conflict) and, admittedly, participate in our enthusiasm about this rewarding collaboration.

We have learned not to see both imaging modalities or their operators as competitors for the same patient, but as adjuncts to one another. ‘Whatever the nature of the technology we use, imaging techniques are tools to solve clinical problems’. There is no need for conflict but only for open and constructive collaboration. In this process we must continue to question the role of both partners and techniques. Fetal MRI is still a new discipline and it is driven by technology that advances today more quickly than does that for ultrasound. In some respects, the debate on fetal MRI is based on a moving target, and the trade-off between different imaging modalities for individual fetal problems will continue to change.


  1. Top of page

M. Cannie*, J. Jani†, S. Dymarkowski*, D. Levine‡, J. Deprest†, * Department of Radiology, University Hospital Gasthuisberg, Leuven, Belgium, † Department of Obstetrics and Gynaecology (Fetal Medicine Unit), University Hospital Gasthuisberg, Leuven, Belgium, ‡ Beth Israel Deaconess Medical Center, Department of Radiology, Boston, USA