Human Reproduction vol.12 no.9 pp.1873–1876, 1997
OUTSTANDING CONTRIBUTION
Assessment of embryonic anatomy at 6–8 weeks of
gestation by intrauterine and transvaginal sonography
Toshiyuki Hata1,3, Atsushi Manabe1,
Ken Makihara1, Kohkichi Hata1, Kohji Miyazaki1
and Daisaku Senoh2
1Department
of Obstetrics and Gynecology, Shimane Medical
University, Izumo 693, and 2Department of Obstetrics and
Gynecology, Tango Central Hospital, Kyoto 627, Japan
3To
whom correspondence should be addressed
Our purpose was to compare the ultrasound visualization
of the early first-trimester embryo using transvaginal and
intrauterine sonography. In all, 32 women about to undergo
therapeutic abortion at 6–8.9 weeks gestation were studied
using a specially developed catheter-based, high-resolution,
real-time miniature (2.4 mm outer diameter) ultrasonography transducer (20 MHz). Before the intrauterine
sonographic procedure was performed, transvaginal sonographic assessment of the embryo was conducted. The
parameters evaluated included the ability to visualize anatomical structures and a subjective assessment of the overall
image clarity. The ability to view most organs was better
with intrauterine sonography compared to transvaginal
sonography, and this was especially true for the brain,
spine, heart, liver, midgut herniation, extremities, and
sacral tail. Moreover, it was possible to obtain finer image
quality of very small embryonic structures with intrauterine
sonography than with transvaginal sonography. Stomach,
spleen, kidney, and bladder could not be depicted with
both techniques. One cystic hygroma was diagnosed at 7
weeks 6 days using intrauterine sonography, but not with
transvaginal sonography. Intrauterine sonography may
provide additional information on the visualization of
anatomical structures of the embryo in the early first
trimester of pregnancy. In this limited series, one case of
cystic hygroma was demonstrated and, thus, there is a
potential for its use in the early detection of embryonic
malformation. These results suggest that intrauterine sonography may be a valuable tool in imaging the early
first-trimester embryo, complementing and not replacing
transvaginal sonography in high-risk pregnancies.
Key words: anatomy/embryo/intrauterine sonography/malformation/transvaginal sonography
Introduction
As a consequence of improvements in ultrasound technology
and with the advent of transvaginal sonography (TVS), visual© European Society for Human Reproduction and Embryology
ization and an appreciation of normal embryonic development
may allow structural anomalies to be detected in the first
trimester of pregnancy (Timor-Tritsch et al., 1990; Cullen
et al., 1990). The clinical utility of TVS is now well established;
however, its ability to depict the anatomy of the early human
embryo is less than satisfactory (Ragavendra et al., 1991).
Sonographic visualization of embryonic anatomy may be
potentially useful in the study of embryonic development and,
possibly, in the detection of structural abnormalities of the
human embryo.
With recent advances in miniaturization of the ultrasound
transducer, Goldberg et al. (1991) showed the feasibility of
passing flexible catheter-based high-resolution real-time ultrasound transducers into the endometrial canal and Fallopian
tube of patients with uterine abnormalities. Potential obstetric
applications of intrauterine sonography (IUS) for systemic
examination of the developmental stages of the early embryo
or detection of gross embryonic malformations have been
reported (Ragavendra et al., 1991, 1993). However, image
quality was poor because the frequency of the transducer used
was 12.5 MHz. In our previous studies (Fujiwaki et al., 1995;
Hata, 1996; Hata et al., 1996), the ultrasonographic frequency
used was 20 MHz, and it was possible to visualize very small
embryonic structures. However, to the best of our knowledge
there has been no report comparing the visualization of
embryonic anatomy by TVS and IUS in the early first trimester.
The aim of this study was to assess embryonic anatomy at 6–
8.9 weeks gestation, comparing the visualization by TVS
and IUS.
Materials and methods
A total of 32 women (five at week 6, 16 at week 7, and 11 at week
8) about to undergo therapeutic abortion at 6–8.9 weeks gestational
age were studied with a specially developed catheter-based highresolution real-time miniature (2.4 mm in outer diameter) ultrasound
transducer (20 MHZ, Aloka AMP-PN20–08L; Aloka Co, Tokyo,
Japan) (Figure 1). The depth of penetration of the ultrasound beam
is ~2 cm. This ultrasonic catheter is connected to an ultrasound
device (Aloka SSD-550). A motor in the main imaging device (Aloka
ASU-100) rotates the metal drive shaft at 900 r.p.m., resulting in a
360° real-time gray-scale image, oriented perpendicularly to the long
axis of the ultrasonic catheter. All examinations were performed by
one examiner (T.H.). The study was approved by the local ethical
committee of Shimane Medical University, Japan, and standardized
informed consent was obtained from each patient.
Before each IUS procedure, an evaluation of the embryo was
performed by TVS (6 MHz Mochida MEU-1581, Tokyo, Japan).
1873
T.Hata et al.
Table I. Percentage of anatomical structures visualized at each gestational
age by transvaginal (TVS) and intrauterine sonography (IUS)
Anatomical structures
Figure 1. Whole view of catheter-based miniature transducer.
Each patient was prepared and draped in the usual sterile fashion in
the dorsolithotomy position. A sterile speculum was inserted into the
vagina. The ultrasonic catheter was introduced gently through the
cervix and into the endometrial cavity until it could not be advanced
any further. Once within the endometrial cavity, the catheter tip was
advanced or withdrawn slightly until the embryo was visualized.
The gestational age by menstrual history was compared with that
calculated from the crown–rump length (CRL) (Iwamoto, 1983).
Only those cases with a discrepancy of ,3 days were included in
the study.
For each gestational age group we recorded the number of cases
in which a structure was adequately evaluated. Correlation of the
detected structures with the appropriate structure as depicted in a
textbook of embryology (Moore, 1982) was attempted for each
gestational age, and the percentage of anatomical structures visualized
at each gestational age was calculated. Visualization of embryonic
organs by TVS and IUS was compared using Fisher’s exact test
(Siegel and Castellan, 1988). P , 0.05 was considered to be
statistically significant.
Results
Three cases (two at week 7, and one at week 8) were excluded
from the study because of the shallow scanning range of highfrequncy transducers or inappropriate embryonal position.
There was no difficulty in passing the imaging catheter through
the cervix into the endometrial cavity. Neither bleeding nor
leakage of amniotic fluid from the external cervical os was
seen after removal of the catheter. There were no known
immediate complications.
Table I itemizes the anatomical structures distinguished at
each gestational age interval by TVS and IUS. The image
clarity of IUS was subjectively superior in all cases studied,
and it was possible to obtain finer image quality of very small
embryonic structures with IUS (Figure 2).
Week 6
One third of the fetal anatomical structures examined (primary
brain vesicle, spine, midgut herniation, upper and lower limbs,
and sacral tail) could be visualized by IUS, but not by TVS.
1874
Primary brain vesicle
Secondary brain vesicle
Spine
Heart movement
Four-chamber view
Stomach
Liver
Spleen
Kidney
Bladder
Midgut herniation
Upper limb
Finger
Lower limb
Toe
Sacral tail
Amniotic membrane
Yolk sac
aP
Gestational age
6–6.9
(n 5 5)
7–7.9
(n 5 14)
8–8.9
(n 5 10)
TVS
IUS
TVS
IUS
TVS
IUS
0
0
0
60
0
0
0
0
0
0
0
0
0
0
0
0
60
100
40
0
60a
100
0
0
0
0
0
0
20
40
0
40
0
40
80
100
71
14
57
100
0
0
7
0
0
0
43
43
0
36
0
14
93
100
50
50a
100
100
21
0
100a
0
0
0
100a
86a
14
86a
14
100a
100
100
50
50
90
100
0
0
80
0
0
0
90
100
0
80
0
90
100
100
10
90
100
100
40a
0
100
0
0
0
100
100
30
100
10
100
100
100
, 0.05, TVS versus IUS at each gestational age.
Heart activity was evident in all cases by IUS, but could not
be recognized in two cases (40%) by TVS.
Week 7
The ability to view most organs was better with IUS compared
with TVS, and this was especially true for the brain, spine,
liver, midgut herniation, upper and lower limbs, and sacral
tail. The proportion of secondary brain vesicles visualized with
IUS was significantly higher than that with TVS (P , 0.05).
Some anatomical structures (four-chamber view, finger, and
toe) could be depicted by IUS, but not by TVS. In one subject
at 7 weeks 6 days a translucent cystic part (3.7 mm in diameter)
behind the nuchal region (Figure 3), which was thought to be
compatible with the ultrasonographic image of cystic hygroma,
was identified by IUS, but not by TVS.
Week 8
The ability to view most anatomical structures by the two
techniques was considered to be equal. However, some structures (four-chamber view, finger, and toe) could be depicted
by IUS, but not by TVS.
Stomach, spleen, kidney, and bladder could not be seen with
both techniques during this period. Amniotic membrane was
depicted in most cases, and yolk sac could be identified in
all cases.
Discussion
The embryonic period, which ranges from week 4 to week 8
after the last menstrual period, is very important for human
development because most major anatomical structures begin
to develop during these 5 weeks. By the end of the embryonic
period most major organ systems have been formed (Moore,
1982). Detailed assessments for the sequential appearance of
Embryonic anatomy by intrauterine sonography
Figure 2. Transvaginal and intrauterine scans of an embryo at 8 weeks 4 days gestation. Transvaginal (TVS) and intrauterine sonographic
(IUS) images in A, B, or C show the same level of the embryo. (A) Oblique plane of embryonic head (H); (B) sagittal plane of embryo (E);
(C) coronal plane of embryo (E). AM 5 amniotic membrane; VD 5 vitelline duct; YS 5 yolk sac, C 5 catheter.
Figure 3. Cystic hygroma (*) at 7 weeks 6 days gestation. (A) Transvaginal scan (TVS); cystic hygroma cannot be identified.
(B) Intrauterine scan (IUS); cystic hygroma (*) is clearly depicted. AM 5 amniotic membrane; E 5 embryo; GS 5 gestational sac; H 5
head; LL 5 lower limb; ST 5 sacral tail; C5 catheter.
embryonic structures by means of TVS have been reported
(Timor-Tritsch et al., 1990; Cullen et al., 1990). Fujiwaki et al.
(1996) reported that IUS could reveal embryonal structures 1
to 3 weeks earlier than could TVS. However, there has been
no report comparing the visualization of embryonic anatomy
by TVS and IUS in the early first trimester. Our study
demonstrates that the ability to view most organs was better
with IUS compared with TVS, and this was especially true
for the brain, spine, heart, liver, midgut herniation, extremities,
and sacral tail. Moreover, the image clarity of IUS was
subjectively superior in all cases studied, and it was possible
to obtain finer image quality of very small embryonic structures
with IUS.
Ragavendra et al. (1991) partly justified their enthusiasm
for IUS by stating that the ability of TVS to depict the anatomy
of the early human embryo was less than satisfactory. Filly
(1991) disagreed with that statement, and made a personal
(and to some extent theoretical) comparison of the two techniques. His comparison demonstrated that IUS had the potential
ability to depict embryonic anatomical detail to a greater
1875
T.Hata et al.
degree than could TVS, yet in virtually every other category
of convenience and utility it was lacking by comparison. With
respect to the common indications for sonography in the first
trimester, IUS appeared to lack the manoeuverability, depth of
penetration, or field of view necessary to permit evaluation of
virtually all common problems. Moreover, Filly (1991) stated
that the ability to allow visualization of heartbeat was exceptionally good with TVS. We also think that IUS lacks the
manoeuverability, deep beam penetration, or ability necessary
to allow perpendicular planes of section to be obtained.
However, as indicated in this study, the ability to depict
anatomical detail of the embryo and image clarity by IUS was
superior to those by TVS in all cases studied. Moreover, heart
activity was evident in all cases by IUS, but could not be
identified in 2 of 5 cases (40%) by TVS at 6 weeks. Therefore,
IUS could be used as an ancillary technique, mainly for
detailed embryonic visualization. However, the activity of
embryonic heart detected by TVS is low compared to other
reports in the literature (Coulam et al., 1996). The reason that
the identification of embryonic cardiac activity at 6 weeks by
TVS in our study is so low is currently unknown. One possible
explanation might be the different detection techniques used
by Coulam et al. (1996) and ourselves for embryonic cardiac
activity. Coulam et al. (1996) used both B-mode and
simultaneous B- and M-modes for detection of embryonic
cardiac activity, while only B-mode was used in the present
study. Another possible explanation is that the size of embryo
may affect the detection of embryonic cardiac activity by TVS.
In the present study, the embryonic pole length in the two
cases in which detection of embryonic cardiac activity by TVS
at 6 weeks failed was 1.3 and 1.6 mm respectively. Hence,
failure to detect the embryonic cardiac activity by TVS in
these cases may be due to the very small embryonic size.
In this study, three cases were excluded because of the
shallow scanning range of high-frequency transducers or inappropriate embryonic position. The depth of penetration of the
ultrasound beam used in our study is ~2 cm, which might be
sufficient to evaluate embryos ,20 mm. However, examination
of a larger embryo or one remote from the transducer is
markedly limited because of the shallow scanning range of
high-frequency transducers. IUS with high-resolution transducers (20 MHz) is suitable for examination prior to week 8,
but not for examination at week 9 (Fujiwaki et al., 1995).
Cystic hygroma is usually diagnosed at 10–12 weeks with
TVS (Cullen et al., 1990). In our previous study (Fujiwaki
et al., 1995) IUS depicted one cystic hygroma at 8 weeks 5
days. In this study, a single case of cystic hygroma at 7 weeks
6 days was diagnosed in utero with IUS, but not with TVS.
Cystic hygromas develop early in gestation at ~day 40 of fetal
development (day 54, menstrual age) (Neiman, 1990), as the
result of non-communication between lymphatic and venous
channels in the neck (Byrne et al., 1984). This may be the
earliest case in which cystic hygroma was diagnosed during
the embryonic period. Unfortunately, pathological examination
and chromosome analysis could not be performed because the
embryo was damaged during the therapeutic abortion.
With respect to the limitations of IUS, it is an invasive
diagnostic procedure requiring sterile conditions and its safety
1876
has not been established. Although we and previous authors
(Ragavendra et al., 1991, 1993; Fujiwaki et al., 1995; Hata
et al., 1996) encountered no immediate complications, the use
of IUS is not recommended for routine clinical examination
in pregnancy at present.
In conclusion, IUS provides additional information on the
visualization of anatomical structures of the embryo in the
early first trimester of pregnancy. In this limited series one
case of cystic hygroma was demonstrated and, thus, there is a
potential for its use in the early detection of embryonic
malformation. These results suggest that IUS may be a valuable
tool in imaging the early first-trimester embryo, complementing, not replacing, TVS in high-risk pregnancies.
References
Byrne, J., Blanc, W.A., Warborton, D. et al. (1984) The significance of cystic
hygroman fetuses. Hum. Pathol., 15, 61–66.
Coulam, C.B., Britten, S. and Soenksen, D.M. (1996) Early (34–56 days
from last menstrual period) ultrasonographic measurements in normal
pregnancies. Hum. Reprod., 11, 1771–1774.
Cullen, M.T., Green, J., Whetham, J. et al. (1990) Transvaginal ultrasonographic detection of congenital anomalies in the first trimester. Am. J. Obstet.
Gynecol., 163, 466–476.
Filly, R.A. (1991) Earlier diagnosis of fetal anomalies: quo vadis? Radiology,
181, 627–628.
Fujiwaki, R., Hata, T., Hata, K. and Kitao, M. (1995) Intrauterine ultrasonographic assessments of embryonic development. Am. J. Obstet. Gynecol.,
173, 1770–1774.
Goldberg, B.B., Liu, J.B., Kuhlman, K. et al. (1991) Endoluminal gynecologic
ultrasound: preliminary results. J. Ultrasound Med., 10, 583–590.
Hata, T. (1996) Intrauterine ultrasonography for the assessment of embryonic
development. Med. Imaging Int., 6, 11–15.
Hata, T., Fujiwaki, R., Senoh, D. and Hata, K. (1996) Intrauterine sonographic
assessments of embryonal liver length. Hum. Reprod., 11, 1278–1281.
Iwamoto, K. (1983) Estimation of gestational age with ultrasonic measurement
of the fetus in each trimester. Acta Obstet. Gynecol. Jap., 35, 2330–2338.
Moore, K.L. (1982) The Developing Human. 2nd edn. W.B.Saunders,
Philadelphia, USA, pp. 227–254.
Neiman, H.L. (1990) Transvaginal ultrasound embryography. Sem. Ultrasound,
CT and MR, 11, 22–33.
Ragavendra, N., McMahon, J.T., Perrella, R.R. et al. (1991) Endoluminal
catheter-assisted transcervical US of the human embryo. Radiology, 181,
779–783.
Ragavendra, N., Beall, M.H., McMahon, J.T. et al. (1993) Transcervical
sonography: an investigational technique for visualization of the embryo.
Obstet. Gynecol., 81, 155–158.
Siegel, S. and Castellan, N.J. (1988) Nonparametric Statistics for the
Behavioral Sciences. 2nd edn. McGraw-Hill Book Company, New York,
pp. 103–111.
Timor-Tritsh, I., Peisner, D.B. and Raju, S. (1990) Sonoembryology: an organoriented approach using a high-frequency vaginal probe. J. Clin. Ultrasound,
18, 286–298.
Received on February 18, 1997; accepted on June 12, 1997