British Journal of Radiology 75 (2002),215-219 © 2002 The British Institute of Radiology
Assessment of fetal anatomy in the first trimester using two- and three-dimensional ultrasound
G D Michailidis
P Papageorgiou
and
D L Economides, MD, FRCOG
Fetal Medicine Unit, Department of Obstetrics and Gynaecology, Royal Free Hospital, Pond Street, London NW3 2QG, UK
Correspondence: Mr D L Economides
 |
Abstract
|
|---|
The objective of this study was to perform a complete anatomical survey of the fetus at 12–13 weeks gestation using stored volumes acquired by a three-dimensional (3D) scanner. 159 consecutive women at 12–13 weeks gestation who had a routine early pregnancy scan in our unit were recruited. A complete survey of the fetal anatomy was attempted by two-dimensional (2D) transabdominal and, if needed, transvaginal ultrasound. Then, using a 3D transvaginal probe, two volumes of the whole fetus were acquired. A complete anatomical survey (excluding anatomy of the heart) was attempted using the stored data. A complete anatomical survey was achieved in 93.7% (149) of cases with 2D ultrasound compared to 80.5% (128) of cases with 3D volume acquisition (p<0.001). The nuchal translucency was measured with 2D scanning in 98.7% of cases and in 91.8% of cases using 3D volumes. The mean time to perform a 2D scan was 12.2 min standard deviation (SD 3.4 min) while the mean time to obtain and examine the stored volumes was 8.4 min (SD 1.45 min, p<0.001). Real-time 2D ultrasound is still the best way to examine fetal anatomy in the first trimester. However, 3D ultrasound can be a useful addition to clinical practice, providing views not easily obtained by conventional 2D ultrasound. It can potentially minimize actual scanning time and provides an excellent way to store scanned data.
|
Introduction
|
|---|
First trimester ultrasound has evolved from simply confirming fetal viability and determining gestational age to an effective screening method for structural and chromosomal abnormalities [1]. Braithwaite et al [2] reported that a complete anatomical survey of the fetus, similar to the criteria of the second trimester scan, can be achieved in 98% of cases at 12–13 weeks gestation. When first trimester ultrasound was used in screening a low risk unselected population, 68% of structural abnormalities and 79% of chromosomal abnormalites were diagnosed at 11–14 weeks gestation [3]. However, to perform the anatomical survey, including the nuchal translucency (NT) measurement, extra scanning time and training is needed [4].
In a series of recent case reports, three-dimensional (3D) ultrasound has been used to help in the diagnosis of structural abnormalities [5] in early pregnancy [6]. The proposed advantages were the ability to reconstruct 3D views, with the scanned volume stored and examined without having to rescan the patient [7, 8]. Although when used in selected cases 3D ultrasound has offered diagnostic advantages, there are no studies comparing the routine use of 3D volume examination with conventional two-dimensional (2D) ultrasound.
The aim of this study was to compare the use of combined real-time 2D transabdominal scanning (TAS) and transvaginal scanning (TVS) with 3D TVS in order to assess the feasibility of performing a complete anatomical survey, including measurement of NT, using stored 3D volumes acquired with minimal scanning time and irrespective of fetal position.
|
Methods
|
|---|
185 women participated in this study. Women with a live singleton pregnancy who attended our routine dating scan clinic at 12 weeks to 13 weeks and 6 days gestation and with a crown–rump length (CRL) of 55–84 mm were included. 26 women were excluded from the study on this basis. All 159 remaining women initially underwent abdominal ultrasound scan to confirm viability and gestational age. An anatomical survey including NT measurement was performed using TAS and, if necessary, TVS according to the criteria presented in Table 1
. For TAS either a Toshiba (Tokyo, Japan) 140 scanner with a 5 MHrz probe, or a Kretz 530D scanner (Kretztechnik, Zipf, Austria) with a 5 MHrz probe, was used. For TVS the Kretz S-VDW 7MHrz probe (Kretztechnik, Zipf, Austria) was used. The transvaginal route was selected for 3D volume acquisition because it provided the best resolution for the reconstructed 3D plane. All scans were performed by two research fellows trained to perform 2D and 3D first trimester scanning. Their experience included 2 years in 2D and 6 months in 3D scanning. The time needed to complete each scan was noted. This study was approved by the hospital's ethics committee and all the women involved gave informed consent.
After the anatomical survey was completed, 3D volume acquisition was performed using the transvaginal 3D probe. The probe uses an integrated motor to produce a slow tilting movement similar to that of a conventional 2D probe, and automatically performs the volume scan. Figure 1 demonstrates how the volume scan was performed. Care was taken so that the fetus was at rest during volume acquisition. Two volumes were obtained. The only requirement for volume acquisition was for the fetus to be at rest and no extra time was allowed for the fetus to change to a more favourable position for scanning. The time needed to obtain the volumes was noted. Volumes were stored in a PC and examined at a later time.
View larger version (55K):
[in this window]
[in a new window]
|
Figure 1. The three-dimensional probe using an integrated motor produces a slow tilting movement similar to that of a conventional two-dimensional probe, and automatically performs the volume scan.
|
|
Using the same criteria, an anatomical survey was performed using the 3D View software installed in one of the unit's PCs and CRL and NT measurements were obtained. The operator performing this 3D reconstruction was blinded to the findings of the 2D scan previously performed by a different operator. The time needed for the volume examination was recorded. The intra and interobserver variability for CRL and NT measurements performed using 3D volumes were examined for the first 30 patients. All 159 women subsequently had a second trimester anomaly scan in our hospital. Results of these were available from the departmental computerized records.
The standard deviation (SD) of the difference between two measurements was calculated to examine the intra and interobserver variability for the first 30 2D and 3D CRL and NT measurements. McNemar's test was used to compare fetal organ visualization by 2D and 3D ultrasound.
|
Results
|
|---|
Of the 185 women who agreed to participate in this study, 26 (14%) had a CRL of either less than 56 mm or more than 84 mm and so were excluded. The mean maternal age of the 159 women studied was 31.1 years (range 18–43years) and the mean CRL was 70.4 mm (range 56–84 mm).
Complete anatomical surveys using 2D scanning were possible in 93.7% (149) (95% confidence interval (CI) 89–97%) of cases. In 75% (112) of these cases the anatomical survey was completed using only TAS. Examining the stored 3D volumes, a full anatomical survey was completed in 80.5% (128) (95% CI 74–87%) of cases. The fetal head, neck, spine, stomach and abdominal wall were more frequently visualized using real-time 2D ultrasound rather than 3D volumes, although differences were not significant. The fetal kidneys were the only organs more often visualized using 3D volumes rather than 2D ultrasound, although the difference was not statistically significant. The fetal feet and bladder were better seen using 2D ultrasound than 3D volumes. Table 2 presents the visualization of fetal organs by each mode.
View this table:
[in this window]
[in a new window]
|
Table 2. The ability to visualize various fetal organs by two-dimensional (2D) real-time ultrasound examination and examination of three-dimensional (3D) stored volumes given as the percentage of cases in which these organs were visualized, and the time needed to aquire images and complete the anatomical survey
|
|
The main reasons for failure to complete the anatomical survey using 3D volumes were fetal position (48.4%; 15 of 31 cases) and fetal movements (25.8%; 8 of 31 cases) during the volume acquisition. The restricted resolution of structures far away from the probe is one of the recognized disadvantages of TVS. This, in combination with the limited manoeuvrability of the transvaginal probe, resulted in sub-optimal 3D images owing to fetal position in 9.4% of all cases (48.4% of the incomplete 3D anatomical surveys). One of the requirements to obtain a 3D volume was for the fetus to be at rest. However, subtle movements, especially of the arms and feet, were occasionally captured resulting in sub-optimal 3D views.
CRL was measured in all cases. The SD of the differences for the intra and interobserver 2D measurements were 2.5 mm and 3 mm, respectively. For the 3D measurements the SD of the differences were 3 mm and 3.5 mm NT was measured in 156 (98.1%) cases using 2D real-time ultrasound, and in 146 (91.8%) of cases by examining the 3D volumes. The SD of the difference for the intra and interobserver measurements was 0.26 mm and 0.2 mm respectively, for the 2D measurements and 0.12 mm and 0.23 mm, respectively, for the 3D measurements.
The mean scanning time for 2D examinations was 12.2 min (SD 3.4 min). The mean time for 3D volume acquisition was 2.9 min (SD 0.8 min). The mean time to examine fetal anatomy and measure CRL and NT using the stored 3D volumes was 5.7 min (SD 1 min). The total time for 3D examination was therefore 8.4 min (SD 1.5 min), significantly less than the 2D scan (p<0.001) (Table 2
).
There were no abnormalities detected at the 20 weeks scan in any of the cases. Three cases with mild hydronephrosis and two with choroid plexus cysts (all detected by both 2D and 3D ultrasound) had resolved by this time. One case with increased NT had chorionic villus sampling and the karyotype was normal.
|
Discussion
|
|---|
This study demonstrates that it is possible to perform a comprehensive anatomical survey of the fetus at 12–13 weeks gestation using a 3D volume in the majority of cases and within a reasonable time frame. Although a complete anatomical survey was more frequently completed using real-time 2D scanning, the examination of 3D volumes, irrespective of fetal position, provided enough information for a comprehensive review of the fetal anatomy in 80.5% (128) of cases [9].
The use of 3D volume examination was evaluated in a routine setting in which 3D volumes, acquired quickly, with minimal actual scanning and irrespective of the fetal position, were used to examine fetal anatomy and measure NT. Fetal head, face, neck, anterior abdominal wall, stomach and spine could be examined using 3D volumes in between 95% and 100% of cases. These structures represent the main sites in which fetal abnormalities can be seen in the first trimester [3]. Using 2D ultrasound, the anatomy of these structures could be examined more frequently, but the difference compared with 3D volumes was not statistically significant. An advantage of 3D volume acquisition is that obtained volumes can be manipulated and, by rotating the image, desired views can be achieved allowing the 3D architecture of these structures to be examined in detail (Figure 2). This is especially useful for the examination of fetal kidneys, the only organs more adequately visualized (p>0.05) by 3D rather than 2D ultrasound. By obtaining a coronal view of the fetal body, the kidneys were easily seen as hyperechogenic structures lateral to the spine in contrast to the hypoechogenic adrenals.
View larger version (94K):
[in this window]
[in a new window]
|
Figure 2. The required views to demonstrate normal fetal anatomy can be obtained from the stored three-dimensional volume.
|
|
3D ultrasound has been used as an alternative method for the accurate measurement of NT. In a recent study Chung et al [10] suggested that the acquisition of 3D volumes can be a very effective and reproducible method of measuring NT. Kurjak et al [11] measured NT in 100% of cases using 3D ultrasound compared with 85% of cases using 2D ultrasound. Even though no extra time was allowed for the fetus to move to a more favourable position for scanning, NT was accurately measured in the majority of cases 92%. The intraobserver variability for measurements taken using 3D ultrasound was less than for measurements taken using 2D ultrasound. In cases where the nuchal region could be clearly seen, NT was easily obtained regardless of fetal position. The image could be rotated in order to obtain a good sagittal view and, by examining the fetus simultaneously in two planes, the amniotic membrane could be distinguished. However, the above advantages were lost when the fetal position did not allow optimal visualization of the fetal neck, e.g. when the fetal head was far away from the transvaginal probe. We believe this is the main reason why the results reported by Chung et al [10] could not be reproduced here, since in that study one of the requirements was the optimal visualization of the nuchal region for acquisition of the 3D volume.
Although the anatomical survey was completed in more cases using conventional 2D ultrasound, the use of 3D volumes resulted in a reduced scanning time required to obtain this information. On average it took only 2.9 min to acquire this volume. The rest of the examination could then be completed without scanning the fetus, making the actual scanning time significantly less than that of conventional 2D ultrasound (12.2 min). The obtained images could then be manipulated in order to get all the desired views. The potential advantages in clinical practice could be reduced examination time as well as providing an excellent way to store ultrasound data, allowing the virtual scan to be repeated at any time.
The use of stored 3D volumes has, however, certain limitations. The loss of the ability to examine the fast moving heart is one disadvantage. In this study the two main factors responsible for a failure to complete the anatomical survey were fetal movements during volume acquisition and limited resolution in certain fetal positions. Fetal movements were responsible, especially at the beginning of our study, for the failure to demonstrate fetal extremities. Although gross movements of the fetal body could easily be seen during volume acquisition, small movements of the limbs were sometimes not recognized, resulting in sub-optimal 3D volume acquisition and artefacts. This problem became less significant as the operators became more familiar with the technique of 3D volume acquisition. Reduced resolution of structures far away from the vaginal probe limited the advantages of 3D manipulation of the obtained volume. Thus, as the 3D reconstructed views could be as good as the respective 2D images, these fetal parts could not be fully examined at the given volume.
Real-time 2D ultrasound is the gold standard in examining fetal anatomy. However, this pilot study shows that 3D volumes can provide maximum information in a very limited scanning time, thus reducing examination time. It also offers a new way of storing and conveying information. 3D volumes in digital format can be sent and examined by a specialist, allowing him/her to perform a virtual scan rather than rescanning the patient. Faster acquisition time and improvement of the resolution of the 3D probe could overcome current limitations and make feasible the routine use of 3D volumes in early pregnancy.
Received for publication May 30, 2001.
Revision received October 16, 2001.
Accepted for publication October 25, 2001.
|
References
|
|---|
-
Economides DL, Braithwaite JM. First trimester ultrasonographic diagnosis of fetal structural abnormalities in a low risk population. Br J Obstet Gynaecol 1998;105:53–7.[Medline]
-
Braithwaite JM, Armstrong MA, Economides DL. Assessment of fetal anatomy at 12 to 13 weeks of gestation by transabdominal and transvaginal sonography. Br J Obstet Gynaecol 1996;103:82–5.[Medline]
-
Whitlow BJ, Chatzipapas IK, Lazanakis ML, Kadir RA, Economides DL. The value of sonography in early pregnancy for the detection of fetal abnormalities in an unselected population. Br J Obstet Gynaecol 1999;106:929–36.[Medline]
-
Braithwaite JM, Kadir RA, Pepera TA, Morris RW, Thompson PJ, Economides DL. Nuchal translucency measurement: training of potential examiners. Ultrasound Obstet Gynecol 1996;8:192–5.[Medline]
-
Bonilla-Musoles F, Raga F, Villalobos A, Blanes J, Osborne NG. First-trimester neck abnormalities: three-dimensional evaluation. J Ultrasound Med 1998;17:419–25.[Abstract]
-
Lai TH, Chang CH, Yu CH, Kuo PL, Chang FM. Prenatal diagnosis of alobar holoprosencephaly by two-dimensional and three-dimensional ultrasound. Prenat Diagn 2000;20:400–3.[Medline]
-
Platt LD, Santulli T Jr, Carlson DE, Greene N, Walla CA. Three-dimensional ultrasonography in obstetrics and gynecology: preliminary experience. Am J Obstet Gynecol 1998;178:1199–206.[Medline]
-
Merz E, Bahlmann F, Weber G, Macchiella D. Three-dimensional ultrasonography in prenatal diagnosis. J Perinat Med 1995;23:213–22.[Medline]
-
Roberts LJ, Bewley S, Mackinson AM, Rodeck CH. First trimester fetal nuchal translucency: problems with screening the general population. 1. Br J Obstet Gynaecol 1995;102:381–5.
-
Chung BL, Kim HJ, Lee KH. The application of three-dimensional ultrasound to nuchal translucency measurement in early pregnancy (10–14 weeks): a preliminary study. Ultrasound Obstet Gynecol 2000;15:122–5.[Medline]
-
Kurjak A, Kupesic S, Ivancic-Kosuta M. Three-dimensional transvaginal ultrasound improves measurement of nuchal translucency. J Perinat Med 1999;27:97–102.[Medline]
This article has been cited by other articles:
|
|
|
|
 
G D Michailidis, P Papageorgiou, R W Morris, and D L Economides
The use of three-dimensional ultrasound for fetal gender determination in the first trimester
Br. J. Radiol.,
July 1, 2003;
76(907):
448 - 451.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
