Clarifying the role of three-dimensional transvaginal sonography in reproductive medicine: an evidenced-based appraisal

Abstract

This overview describes and illustrates the clinical applications of three-dimensional transvaginal sonography in reproductive medicine. Its main applications include assessment of uterine anomalies, intrauterine pathology, tubal patency, polycystic ovaries, ovarian follicular monitoring and endometrial receptivity. It is also useful for detailed evaluation of failed and/or ectopic pregnancy. Three-dimensional color Doppler sonography provides enhanced depiction of uterine, endometrial, and ovarian vascularity.

Figures

Figure 1

Two-dimensional vs three-dimensional transvaginal color Doppler sonography. (a) Diagram (left) and two-dimensional TV-CDS (right) showing arcuate, radial, and spiral vessels in follicular phase. (b) Diagram (left) are two-dimensional TV-CDS (right) showing changes in endometrial vascularity throughout the menstral cycle. During the luteal phase, several spiral vessels are detected. (c) Diagram (top) and multiplanar images (left) and magnified 3D TV-CDS (right) of corpus luteum. The multiplanar image shows the vascular "wreath" surrounding the functioning corpus luteum in the longitudinal (top left), short (axial) (top right) and coronal (bottom right) planes. The combined gray scale and 3D TV-CDS image (both left) which is also magnified and shown as the right depicts the numerous vessels surrounding this corpus luteum.

Figure 2

The techniques of volume calculation. Both figures show the typical multiplanar display of a three-dimensional sonographic dataset of the uterus with the three mutually related orthogonal planes at 90-degrees to one another. The upper left image in both displays represents the longitudinal plane (the A plane), the upper right image the transverse plane (the B plane) and the lower left image the coronal plane (the C plane). Volume calculation can be conducted in either the B or C plane. (a) shows the conventional technique of volume calculation in which a series of 'slices' are taken through the volume of interest whilst the contour is outlined in another plane (the transverse plane in this case). The distance between consecutive slices can be varied according to the degree of change in the surface contour and increased for more complex structures. (b) shows the rotational technique of volume calculation in which the dataset is rotated through 180° about a central axis defined by the application of two callipers. The number of planes available for volume calculation are determined by the rotation step shown in the lower left of the image. Here the 30-degree rotation step has been used and the contour outlined in the coronal or C plane using the manual mode. The resultant three-dimensional model is shown in the lower right of the image.

Figure 3

Three-dimensional 'wire' models of the endometrium. When VOCAL is used to calculate volume a three-dimensional model is generated that may be rotated and examined in its natural orientation. In this model the coronal view is seen on the left clearly demonstrates the fundal defect of an arcuate uterus whilst the actual anteflexion of the uterus may be appreciated in the lateral view on the right.

Figure 4

Three-dimensional power Doppler angiography of the uterine blood supply. A three-dimensional dataset containing power Doppler information has been acquired from the uterus. VOCAL has then been used to define the myometrial-endometrial border and to apply a shell 5 mm outside of that contour to define the sub-endometrium, which can be clearly seen to be more vascular than the endometrium itself.

Figure 5

The 'Histogram'. The three indices of vascularity have been calculated for the endometrial model together with the mean grey value. The Vascularisation Index (VI) reflects the degree of power Doppler information within the model and is considered as a percentage therefore whilst the Flow Index (FI) and Vascularisation Flow Index (VFI) include information on the mean power Doppler signal intensity referenced against a scale from zero to one hundred to indicate the minimal and maximum range accordingly.

Figure 6

Intra-subject variation in endometrial blood flow. The degree of power Doppler information seen within Figure 6a is clearly superior to that seen in Figure 6b and this difference is quantifiable through the 'histogram' facility. These images also serve to demonstrate how vascularity is independent of morphometry and varies throughout different phases of the menstrual cycle as the less vascular homogenous endometrium characteristic of the luteal phase in Figure 6b is of greater volume than that in Figure 6a, which is the same endometrium at the end of follicular phase.

Figure 7

Uterine anomalies. Three-dimensional sonography has become the 'gold standard' investigation for the diagnosis and exclusion of congenital uterine anomalies. Its extremely high sensitivity and specificity relate to its ability to demonstrate the plane coronal perpendicular to the transducer face and in doing so allow visualisation of the fundal contour and comparison of the myometrium with the endometrium throughout the uterine length. Figure 7a shows normal cornna with straight contours at the upper aspect of the cavity in contrast to the characteristic concave contour seen in arcuate uteri (Fig. 7b) and the deeper contour of various length seen in sub-septate uteri (Figs. 7c, 7d & 7e). Any indentation of the fundal contour may also be appreciated in the coronal plane as seen in the multiplanar display in Figure 7f. This uterus had been considered normal with conventional ultrasound, which only provides the longitudinal and transverse views seen in the upper two images, but laparoscopy had demonstrated a bulky uterus with a possible fundal defect and a follow-up three-dimensional ultrasound confirmed the presence of a significant septum. The size of the septal defect can be measured as shown in Figure 7g but may be less important than the remaining length of the cavity shown as a bold dashed line.

Figure 8

Endometrial polyps. Endometrial polyps are best appreciated during the late follicular phase when they can be seen against the plump heterogenous endometrium with its characteristic trilaminar pattern may be seen (Fig. 8a). Polyps may also be seen towards the end of menses however or at the 'down-regulation' scan when many women are still bleeding (see below) with the menstrual fluid acting as an in vivo contrast agent (Fig. 8b). The coronal view may be used to locate the position of intrauterine pathology and confirm the presence of individual smaller polyps suitable often confused as large single fibroids with conventional imaging that are suitable for routine polypectomy rather then hysteroscopic resection (Fig. 8c)

Figure 9

Three-dimensional saline infusion contrast sonography. Two-dimensional ultrasound had shown a persistently thick endometrium despite several weeks of pituitary suppression prior IVF treatment and saline infusion sonography has proven unhelpful simply confirming a thick endometrium despite very low serum oestradiol levels. A three-dimensional acquisition was undertaken therefore both before (Fig. 9a) and during the instillation of saline (Fig. 9b) and subsequently analysed off-line to reduce the amount of saline used and overall time needed for the examination. A large sub-mucosal leiomyoma with a broad base was seen originating from the left posterior aspect of the uterus that almost entirely occupied the cavity (Fig. 9b). This allowed the cancellation of the planned IVF treatment with the patient continuing her 'down-regulation' medication for a further two-months at which stage the fibroid was removed transcervically using a modified operative hysteroscope.

Figure 10

Polycystic ovaries. Three-dimensional sonography facilitates objective assessment of the ovarian stroma, through measurement of its mean grey signal intensity, its vascularity and its volume, which may be calculated by subtracting total follicular volume from total ovarian volume (Fig. 9A). Ovarian blood flow is increased and associated with significantly higher three-dimensional indices of vascularity than ovaries with a normal appearance (Fig. 10b).

Figure 11

Ovarian volume calculation. Ovarian volume may be measured in the same way as endometrial volume through the manual delineation of the ovarian cortex.

Figure 12

Controlled ovarian stimulation. Three-dimensional sonography may be used to facilitate standard measurement of follicular diameter (a). During the latter stages of ovarian stimulation the ovaries may contain many follicles, which become progressively harder to measure reliably between observers, as there are no landmarks to aid orientation. Off-line analysis of stored three-dimensional datasets reduces the time spent with the patient and may potentially improve the number of mature oocytes retrieved by identifying the most appropriate time for oocyte collection. Evidence also suggests that three-dimensional rendering allows demonstration of the cumulus oophorus complex in follicles containing a mature oocyte (b).

Figure 13

Ectopic pregnancy. The spatial orientation of three-dimensional ultrasound and its ability to demonstrate the coronal plane of the uterus in its natural position may help in the diagnosis and exclusion of ectopic pregnancy. This is particularly true in cases of heterotropic pregnancy where an intrauterine pregnancy is associated with an ectopic twin gestation. Spontaneous heterotropic pregnancy is rare but may complicate 0.5% of pregnancies arising from IVF treatment with the ectopic gestation located in the cornua (Fig. 13a) or cervix (Fig. 13b).