Fetal imaging: a brief history of the future


  • Y. Ville

    Corresponding author
    1. Centre Hospitalier Intercommunal de Poissy-St Germain, 10 rue du Champ Gaillard, 78300 Poissy, France
    • Centre Hospitalier Intercommunal de Poissy-St Germain, 10 rue du Champ Gaillard, 78300 Poissy, France
    Search for more papers by this author

original image

The first years of this century probably qualify as being mutational for fetal imaging, with the Journal witnessing the increasing impact of three-dimensional (3D) ultrasonography and the ever-broader development of other types of fetal imaging.

Seize the day

Screening and diagnosis are the two aspects of fetal imaging. They differ mainly in their indications, the level of expertise that they require and, to a certain extent, the complexity of the technology supporting the imaging modality. These largely determine the availability of the investigation, its reproducibility and its accuracy1–3. The number of examinations and the gestational age at which they are performed are driven not only by technical and developmental factors but also by economic and societal considerations as well as legal aspects surrounding termination of pregnancy (TOP)4–6. As a result, screening has moved towards earlier gestations, while diagnostic accuracy is increasing at later gestational ages, when TOP either is not an option or has stopped being relevant to the management of the pregnancy. Ultrasound screening in pregnancy can be seen as the chance to check the largest number of pregnancies by the largest number of operators for simple and reproducible criteria in order to make important choices on the management of pregnancy and delivery. Most established screening programs claim figures of around 40–70% sensitivity (for a 5% false-positive rate) for different conditions, such as congenital heart defects and Down syndrome7, 8. However divergent they might be, both screening and diagnosis demand an ever increasing knowledge of fetal development and mastery of the technology.

Are some women more equal than others?

The 11–14-week scan was created as a model of efficacy and equity8. All reasonably trained operators should be able to offer it to all women at the cost of a commitment to quality assurance, another premiere in the field of prenatal ultrasound, that was still dominated by self-training and craftmanship. As an important part of this examination, identification and measurement of nuchal translucency thickness (NT) is certainly the most cost-effective single screening test for fetal aneuploidies and is that preferred by women9–11. However, although NT measurement has been adopted worldwide, it appears paradoxical that national health systems in developed countries including The Netherlands, the United States, France and the United Kingdom, where it was developed, are still after 15 years, struggling with its implementation. This common failure is attributable to a variety of causes, including inappropriate counseling in pregnancy, bureaucratic or archaic regulations, and indigent ultrasound cover in pregnancy12–16. The widest dissemination of NT measurement has been reported within competitive and largely private initiatives or more recently as a national policy in Denmark, where there is adequate availability and quality of ultrasound for the size and requirements of the population17, 18. At the same time, the gap between research and production of efficient screening for the masses has widened further, with research pushing towards an apparently paradoxical concept of specialized screening in pregnancy which would be available for a few women in a few centers, based on the still relatively young concept of contingency screening19. This type of screening begins with the universal offer of a robust test, such as NT measurement or its combination with maternal serum screening (MSM), that will assign women to one of three subsets—containing less than 5%, 10% or 80–85% of the population, but concealing 80%, 15–20% and less than 5% of the abnormal cases, respectively. The first subgroup is offered a diagnostic procedure and the third subgroup exits the screening, while the intermediate group is offered second-line screening. This can be carried out either in the first trimester, by assessment of fetal nasal bones, frontomaxillary facial angle or Doppler evaluation of the ductus venosus or tricuspid regurgitation, or in the second trimester, with a genetic sonogram or MSM19, 20–23. These second-line screening tests are credited with detection rates as high as 80%, but have a wide range of false-positive rates, varying between 1% and 20%, reflecting both the heterogeneity of the populations studied together with a slow learning curve as well as a degree of subjectivity24. Nevertheless, offering a diagnostic test to 1–20% of this 15% subset of the population would only increase the overall invasive testing rate by 0.1–3%.

Breaking new ground, first-trimester screening is now also being used in the prediction of obstetric problems based on maternal biochemistry and Doppler assessment, with an exponential increase in the number of scientific reports25–27. It is unknown if these scientific breakthroughs will convert into a rational strategy to define pregnancy-associated risks whilst keeping the vast majority of pregnant women away from unnecessary medicalization and planning appropriate care in high-risk cases. However, in the past we have been sufficiently unreasonable as to become aware that if confusion prevails, powerful tools become toys for boys that are not only meaningless to most pregnancies but also iatrogenic in non-certified hands.

3D ultrasound has gradually developed into something more than just a new way of looking at already well-documented fetal structures. Outside a small group of pioneers, promotion of 3D technology was perceived as an expensive and obsessive campaign to demonstrate that poor two-dimensional (2D) images obtained with an inconveniently large transducer could be put together into a grumpy fetal face, that was being presented as a breakthrough in fetal imaging. However, over the last 5 years this ultrasound modality has overcome both technical and cultural challenges. Image quality has gained respect from the most demanding operators and the commercial incentive has become diverse and competitive. At the same time, 3D champions have moved away from pretty faces and volumes and into virtual dissection of the relevant fetal anatomy. This approach is now continuously raising the level of expertise of diagnostic ultrasound and, perhaps more importantly, it is proving to be a remarkable incentive and tool for education.

At the beginning of this century, a multiplanar approach to ultrasound was generally considered eccentric or unnecessary and rarely was this concept seen as either innovative or logical. In the old days registrars were often described as having either surgical or obstetric skills. However, many of us have noticed ultrasound skills burgeoning in year-one trainees and a 5-year follow-up often proves us right. It is no coincidence that ultrasound has gradually moved away from being an exclusive and select apprentiship to become a discipline that can be learned while providing the culture and tools of quality assessment that have long been missing. The frontier between education and research is often subtle, especially because of the critical importance of subjectivity and reproducibility, two extreme and antagonistic components of ultrasound examination. Education is therefore likely to remain the most significant individual contribution each one of us can make to the improvement of perinatal care. The Journal has served this cause and will continue do so.

A thorough review of the performance and utility of 3D/4D examination of the fetal heart28 is timely in the wake of guidelines for fetal echocardiography29 and a didactic journey through spatio-temporal image correlation (STIC)30. It is no coincidence that what was long seen as an outsider with a most uncertain future within the 3D family has become the most spectacular and dynamic field of research and development31, 32. The fetal heart remains the greatest fantasy of the ‘brain’ holding the transducer. STIC and all its by-products provide the best examples of both didactic and antiphobic tools that can help overcome the challenge that fetal echocardiography still presents to most sonologists and sonographers33–37. These educational and psychological dimensions may go slightly over the top, but it is noticeable that the cardiologists have not yet come to recognize the necessity of moving to 3D. It is unlikely that they have remained unaware of these developments and even less so that they could not meet the technical challenge. It is, however, intuitive that they may not need to be educated38.

Whether 3D will lead to important breakthroughs in ultrasound screening remains questionable. Fetal heart examination is likely to represent the most sensitive indicator of the effectiveness of screening policies. Although gestational age is a critical factor for performance, which makes different countries very unequal, a technique that is both highly educational and accessible to telemedicine is more likely to succeed39.

Emulation or competition

The concept of the multiplanar neurosonogram was developed by Ilan Timor-Tritsch in 1996 and has finally reached guideline status40, 41. Pioneers in neurosonography have long recognized that this approach was achievable with 2D ultrasound and was facilitated by 3D technology, while this message was also conveyed convincingly by the rapidly expanding field of fetal magnetic resonance imaging (MRI)42, 43. The relative value of expert fetal neurosonography and MRI is currently under debate. What has become a controversy might at least partly have been fed by a geographical misunderstanding43–45; the quality of the information gathered by either technique is dependent largely on gestational age at examination, which is in turn determined partly by restrictions imposed on TOP. The use of MRI of the fetal brain was developed in the United States, where its application has been confined mainly to second-trimester fetuses with suspected brain malformations46. It is therefore not surprising that some experts in fetal neurosonography claim that the success of MRI is merely a failure of ultrasonography47. In contrast, fetal brain MRI in Europe is performed mainly in the third trimester, as a tool to assess gyration and neuronal migration as well as white-matter damage in high-risk fetuses, or in association with abnormalities diagnosed by ultrasound, including ischemic and hemorrhagic processes48–50. Lurking in the background of this controversy is the fear of some obstetricians of ‘losing’ fetal imaging to radiologists: old tale for a new tool. The status quo with ultrasound has become well accepted, probably because there are relatively few experts in either technique who have gained sufficient knowledge of fetal physiology and mastered recent developments of the technique, while the majority of the others share the vast market of ultrasound screening. However, continuous improvement in the development and performance of fetal MRI is beginning to fill the gaps left by ultrasound to assess fetal organ development and, most importantly, function, while progress in ultrasound technology still aims essentially at improving anatomical imaging. Perinatologists are not (yet!) being trained to perform fetal MRI and pediatric radiologists are therefore gaining increasing recognition in the small world of fetal imaging. At this point in our still young but closer relationship, we should remember the difference between emulation and competition. The former allows interesting hypotheses to be investigated and transformed into meaningful practice while still recognizing the necessarily multidisciplinary approach. The latter often leads to dead-ends, wasting time and energy by ignoring or even undermining the unknown. No other aspect of perinatal medicine has moved as quickly as has fetal imaging, providing pivotal diagnostic but also prognostic and therapeutic tools51–57. This should fill us with pride but also with an even greater perception of our responsibility and duty to pregnant women, their fetuses and children.

The name of our Journal has become a symbol at the crossing of two roads since UOG has given the largest exposure to all imaging techniques and specialties. However, or should I say therefore, I do not think that the Journal needs to change its name. There are many more relevant challenges facing the Journal while reporting cutting-edge applications of all imaging techniques, including the promotion of evidence-based medicine and quality control assessment. I am therefore particularly pleased that the first issue of the New Year launches the publication of serial appraisal of state-of-the-art practice by the Cochrane Collaboration together with commentaries on their relevance for imaging58.