Embryonic Liver Morphology and Morphometry by Magnetic Resonance Microscopic Imaging

Abstract

Embryonic liver has a unique external morphology and quantitative morphometry, based on magnetic resonance imaging data of human embryos from the Kyoto Collection of Human Embryos. Liver morphogenesis is strongly affected by the adjacent organs and tissues. The left ventricle develops to the left medial-caudal side, which results in the formation of a depression at left medial region and a prominence bilaterally at the cranial surface of the liver between Carnegie Stage (CS)17 and CS19. An imprint of the stomach that formed at the dorsal left-medial region of the liver became more marked with development until CS23. A depression induced by the umbilicus formed at the ventral region of the liver between CS16 and CS19. An indentation caused by the right adrenal gland formed at the dorsal-caudal region of the liver surface from CS20. Morphometric analysis revealed that the volume of the liver increased exponentially from CS14 through CS23. The liver developed preferentially along the dorsoventral axis and right/left axis until CS17, along the craniocaudal axis between CS17 and CS19, and then in all directions after CS19. Several important developmental phenomena, such as differentiation of the diaphragm, the extension of the body axis of the embryo, and the physiologic herniation of the intestine into the umbilical cord, may affect morphometric data. These data contribute to a better understanding of liver development as well as the morphogenesis of adjacent organs, both temporally and spatially, and serve as a useful reference for fetal medicine and prenatal diagnosis. Anat Rec, 2012. © 2011 Wiley Periodicals, Inc.

The liver occupies a large space in the abdominal cavity during most of the prenatal period and plays an important role in the development of functional organs (Lemaigre, 2009; Sadler and Langman, 2010). The liver becomes a hematopoietic organ after 6 weeks (Drews, 1995) and begins to metabolize important biochemical materials for development, such as albumin, bile, glycogen, and fetal-specific proteins, at around 8 weeks (Carlson, 2009).

The development of the liver proceeds in a unique manner. The liver develops at Carnegie Stage (CS) 11 (30 days after fertilization) as an outgrowth of the endodermal epithelium, the liver bud, from the caudal part of the foregut. The liver originates from two different tissues: angioblastic tissue from the coelomic surface cells and epithelial columns sprouting from the hepatic evagination of the gut epithelium (O'Rahilly and Müller, 1987). The liver lies at an active center of angiogenesis in the early embryonic period. The asymmetricity of the afferent venous vessels of the liver derives from two specific circulation systems: the vitelline and umbilical veins, which are acquired between CS13 and CS16 (Mall, 1906; Dickson, 1957; Collardeau-Frachon and Scoazec, 2008). Efferent venous vessels, including the right, left, and middle hepatic veins (HVs) and the inferior vena cava (IVC), form at similar stages. The developmental process of the efferent venous vessels is not as well studied as that of the afferent venous vessels (Mall, 1906; Dickson, 1957; Couinaud, 1996; Collardeau-Frachon and Scoazec, 2008).

Among recent three-dimensional (3D) imaging techniques, magnetic resonance (MR) microscopy is a powerful tool for 3D measurements. It is a noninvasive and nondestructive imaging method, and has been applied to analyze embryonic development in different animal models (Bone et al., 1986; Smith et al., 1992, 1994, 1996). MR imaging of embryos is highly advantageous (Effmann et al., 1988; Smith et al., 1992; Haishi et al., 2001), providing a resolution of 40 μm/pixel or better with long scan times. Kyoto and Tsukuba Universities began a project in 1999 to acquire 3D MR microscopic images of thousands of human embryos using a super-parallel MR microscope operated at 2.34T (Shiota 2007; Matsuda et al., 2003, 2007; Yamada et al., 2006).

In the present study, the precise external morphology and morphometry of the embryonic liver was studied using MR imaging data of human embryos from the Kyoto Collection of Human Embryos (http://bird.cac.med.kyoto-u.ac.jp). These data will serve as a useful reference for evaluating the development of the embryonic liver and adjacent organs and how they morphologically affect each other.

Abbreviations used: 3D = three-dimensional; CS = Carnegie stages; HV = hepatic veins; IVC = inferior vena cava; L_CC = craniocaudal length; L_DV = dorsoventral length; L_TH = trunk height; L_TR = transverse length; MR = magnetic resonance; ROI = region of interest.

MATERIALS AND METHODS

Human Embryo Specimens

Approximately 44,000 human embryos, comprising the “Kyoto collection,” are historical specimens collected and stored at the Congenital Anomaly Research Center of Kyoto University (Nishimura et al., 1968; Nishimura, 1975; Shiota, 1991; Yamada et al., 2004). In most cases, pregnancy was terminated during the first trimester of pregnancy for socioeconomic reasons under the Maternity Protection Law of Japan. Some of the specimens (∼20%) are undamaged, well-preserved embryos. When the aborted materials were brought to the laboratory, the embryos were measured, examined, and staged using the criteria of O'Rahilly and Müller (1987). Approximately 1,200 well-preserved human embryos diagnosed as externally normal at CS13 to CS23 were selected for MR microscopic imaging. The conditions used to acquire the MR images of the embryos are described elsewhere (Shiota, 2007; Matsuda et al., 2003, 2007; Yamada et al., 2010).

MR Image Processing and Selection of the Datasets

3D MR image datasets for each embryo were initially obtained from 256 × 256 × 512 voxels. Each dataset was first converted into a two-dimensional (2D) stack and saved as an audio video interleave (.avi) file format using software ImageJ™ (version1.42q, National Institutes of Health, Bethesda, MD). Sequential 2D images were resectioned digitally and 3D images were reconstructed using the software OsiriX™ (version 3.7.1, Pixmeo SARL, Geneva, Switzerland). Both 2D and 3D images were carefully observed and selected according the following conditions: (1) no obvious damage or significant anomaly present in the external appearance, (2) body axes maintained in the original form that is not deformed artificially during fixation and preservation, (3) sufficiently high quality of reconstructed 2D images to properly extract the organs and tissues, and (4) liver, stomach, IVC located in the normal anatomic position.

For the present study, 67 samples were selected from the 1,200 MR image datasets based on the criteria described earlier, consisting of five cases each for CS14, CS16, CS18, CS19, CS21, and CS22; nine cases each for CS15, CS20, and CS23; and 10 cases for CS17.

3D Reconstruction of the Liver and Adjacent Organs

The avi file format images obtained from 3D MR images were resectioned digitally to the suitable planes for each analysis using the multiplanar reformatting tool in software Image J and OsiriX. Targeted organs of interest were segmented in a series of coronal-sectional images using the region of interest (ROI) module in OsiriX. 3D objects were computationally reconstructed with DeltaViewer™ (http://delta.math.sci.osaka-u.ac.jp/DeltaViewer/index-j.html; Yamada et al., 2007). Morphogenesis of the liver and adjacent organs was analyzed in detail using the 3D images (see Supporting Information Video S1, S2, S3).

Measurements of the Liver

For accurate anatomic assignment of the cranial/caudal (Z), left/right (X), and dorsal/ventral (Y) axes, the cranial/caudal axis of the liver was first determined using OsiriX. The z-axis was defined as the line that goes through the most cranial and caudal neural tubes in a series of coronal-section images that include the liver (Fig. 1A). The orthogonal plane of the z-axis was then determined and the x- and y-axes were defined as shown in Fig. 1B. Lengths of the liver along each axis were defined as the transverse length (L_TR), dorsoventral length (L_DV), and craniocaudal length (L_CC) (Fig. 1A,B).

Figure 1. Morphometry of the liver (A,B) An orthogonal coordinate system was defined using craniocaudal (Z)-axis, left/right (X)-axis, and dorsoventral (Y)-axis as described in the Materials and Methods section. Length of the liver along each axis was defined as craniocaudal length (L_CC) transverse length (L_TR), and dorsoventral length (L_DV), respectively. Li; liver (C) Trunk height (L_TH), the length between the axilla and the cranial end of the trochanterion (•),was used to measure the change in the crania-caudal growth of the abdominal cavity using the method of Otani et al. (2008) with some modification.

The liver was extracted from a series of coronal-section images as described above. The volume of the liver was calculated by stacking the extracted liver in a series of coronal-section images using the ROI module of OsiriX. The volume of the embryo was calculated using the Region Growing module in OsiriX.

We measured trunk height (L_TH) as the change in the craniocaudal growth of the abdominal cavity according to a previous study by Otani et al. (2008) with some modification (Fig. 1C). In the present article, the axilla was marked instead of the acromion used by Otani et al. (2008), because the axilla was more evident on our 3D images.

Estimation of Vascular Architecture in the Liver

To elucidate the asymmetry of the afferent venous vessels, the following four vessels were reconstructed three-dimensionally; the ductus venosus, umbilical vein, portal vein, and common HV (Fig. 2A). The common HV becomes the intrahepatic part of the IVC in later stages. The three efferent venous vessels, the right, left, and middle HVs (Fig. 2B), were also estimated on 2D serially sectioned images.

Figure 2. Schematic representation of embryonic circulation of the liver. (A) Definitive afferent venous circulation of the asymmetrical stage (B) Definitive efferent venous vessels. Arrow indicates the direction of venous flow. Li, liver; DV, ductus venosus; PV, portal vein; UV, umbilical vein; CHV, common hepatic vein; IVC, inferior vena cava; RHV, right hepatic vein; LHV, left hepatic vein; MHV, middle hepatic vein.

RESULTS

Anatomic Relationships with the Liver

The morphogenesis of the liver was affected by the development of adjacent organs and tissues, such as the heart, diaphragm, stomach, umbilicus, abdominal wall, and adrenal gland. To elucidate the characteristic changes occurring stage- and organ-specifically, the morphogenesis of the liver is described in detail in relation to the development of adjacent organs and tissues.

Intrathoracic Organs

The right and left ventricles were at the same level along the cranial/caudal axis, and the liver was in contact with the ventricles at CS15 in the 2D image. As a consequence, the liver formed a prominence on the bilateral cranial region in all 11 cases at CS15 and CS16 (Fig. 3A-a,b). From CS17, the left ventricle developed to the left medial-caudal side, which resulted in the formation of an obvious depression in the left medial cranial region and prominences bilaterally on the cranial surface of the liver in all 11 cases (100%) between CS17 and CS19 (Fig. 3A-c,d). This depression in the left medial cranial region is a characteristic temporal feature of the liver between CS17 and 19, and is hence termed the “heart depression.” The “heart depression” was deep (maximum) until CS18, and then disappeared in 4 of 8 cases (50%) at CS20 and in 17 of 18 cases (94.4%) after CS21 (Fig. 3A-e,f). Next, the liver formed a prominence in the central region of the cranial surface in all 5 cases at CS22, as the thickness of the diaphragm was sufficiently developed to create a distinct border between the thoracic and abdominal cavities (Fig. 3B-a). The top of this prominence moved toward to the right in 7 of 8 cases (87.5%) at CS23, while the left ventricle developed left-ventrally (Fig. 3B-b).

Figure 3. Representative 3D image of the embryonic liver, demonstrating the anatomic relationship between the intrathoracic organs and the liver. (A) Ventral and left lateral view of the liver by 3D image between CS15 and CS20. The liver (green), lung (blue), and heart (red) were reconstructed in the picture (b, d, f). The depression formed by the left ventricle (*) is a characteristic temporal feature of the cranial surface of the liver between CS17 and CS19. (B) Ventral view of the liver by 3D image at CS22 and CS23. A prominence was formed in the central region of the cranial surface of the liver at CS22. The top of this prominence moved toward the right at CS23. The prominence formed by the left ventricle is a characteristic temporal feature of the cranial region of the liver at CS22 and CS23.

The lung was recognized from CS14, and was in close contact with the dorsal side of the liver between CS14 and CS18 (Fig. 3A-b,d). The lung and liver were clearly separated by the developing diaphragm after CS19 in all 31 cases (Fig. 3A-f). The lung did not seem to affect the morphogenesis of the liver during development.

Stomach

The liver was deformed on the dorsal and caudal surface because of the organogenesis of the upper digestive tract, especially the stomach. The stomach formed hollows on the left-medial regions of the dorsal surface of the liver, that developed a fusiform in all 6 cases at CS15 (Fig. 4A-a,b). Then, the liver developed by covering the stomach along the greater curvature, while the stomach formed a greater and lesser curvature and rotated around in a 3D manner. As a consequence, the oral side of the stomach formed hollows on the dorsal-caudal surface of the liver in all 5 cases at CS18 (Fig. 4A-c,d). As the stomach rotated, the anal side of the stomach formed a loop, which became the pyloric antrum. The liver formed a “horizontal plane” by the loop on the caudal surface after CS19 in 27 of 29 cases (Fig. 4B-a,b). Marked hollows formed on the liver with development, such as the “imprint of the stomach” on the caudal surface in all 8 cases at CS23 (Fig. 4A-e,f).

Figure 4. Representative 3D image of the embryonic liver, demonstrating the anatomic relationship between the stomach and liver. (A) Dorsal view of the liver by 3D image between CS16 and CS23 The liver (green) and stomach (orange) was reconstructed (b, d, f). The “Imprint of the stomach” (*) changed according to the morphogenesis of the stomach and become deeper until CS23. (B) Caudal view of the liver by 3D image showing the “imprint of the stomach” (*) and “horizontal plane by pyloric-antrum” (**).

Umbilicus

The part of the liver is in contact with the abdominal wall curved smoothly. The liver looked like a sector from the lateral view in 5 of 6 cases (83.3%) at CS15 (Fig. 5A-a). The liver developed along the cranial-caudal axis, and a depression formed on the ventral-medial region in all 8 cases between CS17 and CS18 (Fig. 5A-b). This depression results from the entry of the umbilical vein and intestinal tract, which herniates physiologically into the umbilical cord (Fig. 5B-a,b). The depression caused by the umbilicus disappeared in all 26 cases after CS20 because the umbilical cord moved toward the caudal side of the abdomen. Flexure of the abdomen of the embryo was evident after CS20. As a consequence, the liver looked like a quadrangle from the lateral view (Fig. 5A-c).

Figure 5. Anatomic relationship between the umbilicus and liver (A) Representative right lateral view of the liver by 3D image between CS15 and CS20 showing the “depression by umbilicus” (*). This depression is a characteristic temporal feature at the abdominal region of the liver between CS16 and CS19. (B) Lateral view of the liver by 3D image at CS17 showing the relationship between the liver (Li; green) and umbilicus (U; yellow). External form of umbilicus, leg, and chest-abdominal wall is represented by the yellow and dashed lines, respectively. Red and blue circles indicate the entrances to the umbilical vein and intestinal tract, respectively.

Retroperitoneal Organs

Retroperitoneal organs were located close to the dorsal-caudal side of the liver, but only the right adrenal gland directly contacted the liver. The adrenal gland developed remarkably up to CS19, with the right adrenal gland forming an indentation in the liver (Fig. 6A,B) in 3 of 5 cases (60%) at CS19 and in all 26 cases after CS20 in the dorsal-caudal region. The indentation formed by the adrenal gland on the liver was unilateral, only on the right side, mainly because the left adrenal gland was separated from the liver by the stomach (Fig. 6A,B). The metanephros and gonads were recognized on MR images after CS21, and seemed to contact the liver. The metanephros and gonads, however, did not seem to affect the morphogenesis of the liver in any of the 18 cases after CS21 (Fig. 6A,B).

Figure 6. Dorsal view of the 3D image of the liver at CS22 showing the anatomic relationship between the retroperitoneal organs and the liver. The liver (green), bilateral adrenal gland (purple), metanephros (yellow), and the gonads (brown) were reconstructed (B). The right adrenal gland impinged on the liver, creating an indentation (*), while the left adrenal gland remained separated from the liver by the stomach between them.

Morphometry of the Liver

At CS14, the mean volume of the liver was 0.85 ± 0.32 mm³ (mean ± SD) and by CS23 it had increased to 77.40 ± 31.30 mm³ (Fig. 7A). The mean volume of the whole embryo at CS14 was 29.90 ± 7.53 mm³, and at CS23 it had reached 1458.40 ± 433.60 mm³. The ratio of liver volume to whole embryonic volume was 2.8 ± 0.8% at CS14, gradually increasing to 5.7 ± 0.5% at CS22 (Fig. 7B).

Figure 7. Calculated liver volume of embryo from CS14 to CS23 (A) Liver volume was calculated as described in Materials and Methods section. Data at each CS is shown as mean ± SD mm³ (B) Ratio of liver and embryo (vol/vol %). Whole embryonic volume was calculated as described in the Materials and Methods section. Data at each CS are shown as mean ± SD (%).

The length of the liver along the three axes increased exponentially, as shown in Fig. 8A. To clarify the direction of the growth of liver, the ratio of L_TR and L_DV to L_CC according to CS was calculated. The data revealed that the direction of the increase changed at around CS17 and CS19 (Fig. 8B). That is, the liver developed preferentially along the dorso/ventral axis and right/left axis until CS17, along the cranio/caudal axis between CS17 and CS19, and then in all directions after CS19.

Figure 8. Morphometry indicating the 3D direction of liver growth (A) Change in liver length measured along three axes between CS14 and CS23 (B) Ratio of transverse and dorsoventral length (L_TR, L_DV) to coronal craniocaudal length (L_CC) between CS14 and CS23.

L_TH increased from 1.84 ± 0.34 mm at CS14 to 8.68 ± 1.40 mm at CS 23 (Fig. 9A). The ratio of L_CC to L_TH was around 50% between CS14 and CS17, and then it increased to around 75% between CS18 and CS23 (Fig. 9B). The data indicated that the liver occupied about half of the abdominal cavity at CS14 and CS17, and about three-fourths of the abdominal cavity after CS18.

Figure 9. Trunk height (L_TH) (A), and ratio of L_TH and craniocaudal length (L_CC) (B) during CS14 and CS23 embryos Trunk height (L_TH) was measured as a change in the crania-caudal growth of the abdominal cavity as described in the Materials and Methods section.

Vascular Architecture of the Liver

Of 62 cases from CS14 to CS23, 61 were defined as asymmetrical. The umbilical vein, portal vein, ductus venosus, and IVC were recognized in each embryo. Primary right and left HVs emerging symmetrically were identified in only one exceptional case.

The arrangement of the three terminal HVs was examined in 62 cases between CS14 and CS23 (Table 1). No HVs were identified in any of the 5 cases at CS14. All the right, left, and medial HVs were recognized between CS15 and CS23 in 39 of 57 cases (68.4%). The remaining 18 cases varied as follows: right and left HV (8 cases, 14.0%), right HV alone (1 case, 1.8%), left and middle HV (5 cases, 8.7%), left HV alone (3 cases, 5.3%), and no HV (1 case, 1.8%).

Table 1. Acquisition of three hepatic veins from carnegie stage 14–23

Terminal hepatic vein			Carnegie stage										No. of total cases
Right	Left	Middle	14	15	16	17	18	19	20	21	22	23
Acquisition of right, left, and middle hepatic veins is indicated as “y”. Five of 67 cases were unfit for estimation of vascular architecture.
y	y	y	0	1	2	6	4	5	7	4	4	6	39
y	y	–	0	5	2	0	1	0	0	0	0	0	8
y	–	–	0	0	0	0	0	0	0	0	0	1	1
–	y	y	0	0	1	1	0	0	1	1	1	0	5
–	y	–	0	1	0	1	0	0	0	0	0	1	3
–	–	y	0	0	0	0	0	0	0	0	0	0	0
–	–	–	5	1	0	0	0	0	0	0	0	0	6
No. of total cases			5	8	5	8	5	5	8	5	5	8	62

DISCUSSION

The liver bud grows rapidly, and the embryonic liver occupies most of the abdominal cavity after the end of the 6th gestational week (ca. CS16; Hutchins and Moore, 1988; Lemaigre, 2009). The details of the morphologic and morphometric features of the liver during the early embryonic period, however, have remained unknown. Mall (1906) made wax models of the liver exterior from serial histologic sections of the human embryo to study the positional relationship of the gall bladder and vascular system from an outside view. He only described the morphologic changes of the liver in embryos between 17.5 and 24 mm in size; “By comparing the livers of three embryos it is seen that only their upper surfaces are regular in form from stage to stage; the processes extending into the abdominal cavity are irregular, to fit into the spaces that there are for them to grow into.” Severn (1971) examined serial histologic sections of 38 human embryos from CS9 through CS11. For his detailed histologic observation, 3D drawings of the developing foregut and hepatic diverticulum were made, showing the change in the external appearance. Hutchins and Moore (1988) reported that the liver appeared at CS11 and grew to over 90 mm³ in volume by CS23. They calculated the difference in the volume between right- and left-halves of the liver, divided by the median sagittal plane; the right half was large with an average proportion of 57.8%, and the ratio was almost constant in all embryos from CS11 through CS23. In the present study, external morphologic and morphometric analysis of the liver during embryonic periods was performed using MR imaging data acquired from embryos obtained from the Kyoto Collection. The present data revealed a unique external morphology as well as the quantitative morphometry of the embryonic liver.

Morphogenesis of the liver was strongly affected by the adjacent organs and tissues. The characteristic effects of stage- and organ-specific changes are summarized in Fig. 10. The left ventricle developed to the left medial-caudal side, which resulted in the formation of a clear depression in the left medial region and prominence bilaterally on the cranial surface of the liver between CS17 and CS19 (Fig. 3A-c,d). An imprint of the stomach formed at the dorsal left-medial region of the liver, and became more marked with development until CS23 (Fig. 4A-a,c,e). A depression caused by the umbilicus formed in the ventral region of the liver between CS16 and CS19 (Fig. 5A-b). An indentation created by the right adrenal gland was formed at the dorsal-caudal region of the liver surface from CS20 (Fig. 6-a,b). Therefore, the morphology of the embryonic liver reflects the development of the adjacent organs during organogenesis.

Figure 10. Development of adjacent organs and tissues that may affect liver morphogenesis. Temporal (stage-specific) and organ-specific effects are indicated and named according to the Carnegie stage.

Morphometric analysis in the present study revealed that the volume of the liver increased exponentially from CS14 through CS23, and the ratio of L_TR, L_DV, and L_TH to L_CC presented here indicated that the direction of growth changed at around CS17 and CS19 (Figs. 8B, 9B). That is, the liver developed preferentially along the dorso/ventral axis and right/left axis until CS17, along the cranio/caudal axis between CS17 and CS19, and then in all three directions. The occurrence of several important developmental phenomena around CS17 may affect the morphometric data (O'Rahilly and Müller, 1987; Moore, 2008; Schoenwolf and Larsen, 2009). When the septum transversum begins to differentiate into the diaphragm, development in the cranial direction is limited, while development towards the abdominal cavity is accelerated likely due to extension of the body axis of the embryo and physiologic herniation of the intestine into the umbilical cord, which creates space and transform the inner structures of the abdominal cavity (O'Rahilly and Müller, 1987).

Original and first-hand data regarding the stages of development of the vascular architecture of the liver are scarce (Collardeau-Frachon and Scoazec, 2008). Though the asymmetry of the hepatic vascular structure may be acquired between CS13 and CS16, the precise stages of development could not be determined. The right umbilical vein, which is an important indicator of the symmetrical stage, was clearly detected at CS13 (O'Rahilly and Müller, 1987). In the present study, only one case showed the right umbilical vein at CS14 and other 61 of 62 cases had already lost the right umbilical vein by CS14; that is, the hepatic vascular structure was already asymmetrical. The present data suggest that the fundamental architecture of the asymmetrical stage is acquired between CS13 and CS14 in almost embryos.

The terminal HVs formed at a similar stage as the afferent circulation system, as mentioned earlier. In the present study, three HVs were observed in 68.4% of the cases after CS15, indicating that the three HVs are acquired around CS15 in most cases. These data are consistent with those of a previous study reporting profound remodeling of the efferent venous system during the 5th gestational week (Dickson, 1957; Collardeau-Frachon and Scoazec, 2008). Three HVs were not identified until after CS17 in 19.6% of cases, suggesting that there are several individual variations in the number and arrangement of the terminal HVs, in contrast to the afferent venous circulation systems. It is so far impossible to distinguish an anomaly from a variation in individual embryos, mainly because only terminal HVs were detected on the MR image. Detailed identification of such a small branch of the vessels depends on the resolution of the imaging technique. Further improvements in imaging modalities are expected that will allow for more precise detection of the intrahepatic vascular system and application to analyses at CS13 or earlier.

Recent advances in medical imaging allow for earlier assessment of human development and prenatal diagnosis in the first trimester. Data about normal development during the embryonic stages, however, remain inadequate for guiding such clinical evaluations. Insights into the dynamic and complex processes during organogenesis will require accurate morphologic data with dynamic modeling of embryonic structures. Furthermore, 3D reconstructions are necessary to elucidate the complex anatomic remodeling that occurs during these early embryonic stages. From this point of view, the present data will be useful for evaluating the appropriate development of the embryonic liver based on the external morphology, and for evaluating adjacent organs that affect the morphology of the liver stage-specifically. This information will be an indispensable reference for clinical evaluation with obstetrical ultrasonography in the early gestational weeks, which will be useful for fetal medicine and prenatal diagnosis.

Acknowledgements

We are deeply indebted to Executive Vice President of Kyoto University, Kohei Shiota, for providing the invaluable MR data. We also acknowledge the contribution of collaborating obstetricians and the previous members of the Congenital Anomaly Research Center, Kyoto University Graduate School of Medicine.