Automated three-dimensional coded contrast imaging hysterosalpingo-contrast sonography: feasibility in office tubal patency testing


Correspondence to: Dr C. Exacoustos, Obstetrics and Gynecology Department, Università degli Studi di Roma ‘Tor Vergata’, Ospedale Generale S. Giovanni Calibita ‘Fatebenefratelli’, Isola Tiberina 1, 00186 Rome, Italy (e-mail:



To evaluate the feasibility of transvaginal hysterosalpingo-contrast sonography (HyCoSy) with new automated three-dimensional coded contrast imaging (3D-CCI) software in the evaluation of tubal patency and visualization of tubal course.


Patients undergoing HyCoSy with automated 3D-CCI software were evaluated prospectively. First, to evaluate the feasibility of 3D visualization of tubal course, we performed consecutive volume acquisitions while injecting SonoVue contrast agent. We then performed conventional two-dimensional (2D) real-time HyCoSy to confirm tubal patency status by detection of saline and air bubbles moving through the Fallopian tubes and around the ovaries. We also evaluated visualization with CCI of the contrast agent around the ovaries, side effects and pain during and after the procedure, by visual analog scale (VAS) (ranging from 0 to 10, with 0 corresponding to no pain and 10 corresponding to maximum pain).


A total of 126 patients (252 tubes) underwent 3D-CCI HyCoSy followed by 2D real-time HyCoSy. According to the final 2D real-time evaluation, bilateral tubal patency was observed in 111 patients, bilateral tubal occlusion in four patients and unilateral tubal patency in 11 patients. The concordance rate for tubal patency status between the first 3D volume acquisition and the final 2D real-time evaluation was 84% and that between the second 3D volume acquisition and the final 2D real-time evaluation was 97%. A pain score >5 on VAS was recorded in 58% of patients during the procedure, but a pain score ≤ 5 was recorded in 85.7% of patients immediately after the procedure.


HyCoSy with automated 3D-CCI technology retains the advantages of conventional 2D HyCoSy while overcoming the disadvantages. 2D HyCoSy is highly observer-dependent and is only accurate in the hands of experienced investigators; by obtaining a volume of the uterus and tubes, automated 3D volume acquisition permits visualization of the tubes in the coronal view and of the tubal course in 3D space, and should allow less experienced operators to evaluate tubal patency status relatively easily. Copyright © 2012 ISUOG. Published by John Wiley & Sons, Ltd.


Evaluation of Fallopian tubal patency is one of the initial steps in the diagnostic workup of infertile women. Hysterosalpingo-contrast sonography (HyCoSy) is currently performed as an office procedure to assess tubal patency by transvaginal sonography (TVS)[1, 2]. Because it can be performed in an outpatient setting, it is a useful tool in one-stop infertility clinics[3].

HyCoSy is based on the introduction into the uterine cavity and Fallopian tubes of a sonographic-enhancing positive-contrast fluid that can be visualized by TVS. In a previous paper[4] we showed that HyCoSy using saline solution mixed with air is an effective and inexpensive screening tool with which to assess tubal patency status, with high diagnostic accuracy when compared to hysterosalpingography (HSG) and laparoscopic dye perturbation, which is the gold standard procedure in tubal patency evaluation.

However, HyCoSy has limitations. It is highly observer-dependent and it is only accurate in the hands of experienced investigators[4-7], because the Fallopian tubal course is not linear and lies on different planes, so rapid movements of the probe are needed to visualize the entire tubal course and the passage of fluid through the tube during infusion. Furthermore, HyCoSy is not as accurate when tubes are occluded, possibly due to intermittent tubal spasm and to difficulties in distinguishing saline and air in the tubes from air moving in the bowels. Moreover, HyCoSy with air/saline solution does not provide an image of the entire tube and its course, as HSG does.

The use of ultrasound-enhancing contrast media has been introduced to improve the evaluation of tubal patency status[8, 9]. Recently, second-generation contrast agents have been proposed to evaluate various gynecological lesions[10-15] and for sonographic tubal patency evaluation[16-19]. These agents provide a substantial harmonic response at low acoustic pressure and can be seen more easily and endure longer than do the earlier contrast agents. Their safety for intravenous use has been confirmed by the United States Food and Drug Administration and by the European Commission.

Special ultrasound technologies have been proposed[15, 17-19] for optimization of the use of ultrasound contrast media by means of low acoustic pressure. In coded contrast imaging (CCI)[20], the ultrasound machine emits a beam at a selected frequency and receives a narrow band of harmonic signal, avoiding overlap between the tissue and the contrast response. The image displayed is based only on harmonic signals produced by contrast-medium microspheres; broadband ultrasonic signals from surrounding tissue are filtered out completely.

In an attempt to overcome the limitations of conventional two-dimensional (2D) HyCoSy, with its need for an experienced sonographer to visualize the entire tubal course, two ultrasound technologies have been combined: CCI, to better evaluate the signals coming from the contrast medium, and three-dimensional (3D) sonography, to acquire a volume of the Fallopian tubes[20].

Several studies[4-9, 16-20] have shown high accuracy (85–95%) in determining tubal patency with different ultrasound techniques and contrast media, including automated 3D-CCI HyCoSy[21]. In this study, we investigated further the feasibility of HyCoSy with the new automated 3D-CCI software in the evaluation of tubal patency and visualization of the tubal course.


Patients with primary and secondary infertility undergoing evaluation of tubal patency were included in this study. Infertile patients were defined as women who had been actively trying to become pregnant for more than 1 year.

Before the HyCoSy procedure, all patients gave their history and underwent clinical examination including 2D- and 3D-TVS. During 2D TVS we described precisely and recorded the presence of: fibroids; any intrauterine pathology; any adnexal lesion; and retroperitoneal deep infiltrating endometriosis (DIE). We excluded all patients with ongoing pregnancy, reproductive tract cancer or with risk factors such as heart disease, especially heart shunt hypertension and stroke. Exclusion criteria also included presence of pelvic or vaginal infection and tubal pathology that was detectable by TVS (hydrosalpinx, acute salpingitis). Patients who met our criteria underwent 3D-CCI HyCoSy followed by 2D real-time HyCoSy during the proliferative phase of their cycle (Days 5–12). Our Institutional Review Board approved this study, and written informed consent was obtained from each patient.

HyCoSy was performed, using a Voluson E8 ultrasound machine (GE Healthcare, Zipf, Austria) with automated 3D-CCI HyCoSy software (GE Healthcare) and SonoVue (Bracco International BV, Amsterdam, The Netherlands) ultrasound contrast medium, according to the technique described in our previous studies[4, 17, 20].

After inserting a speculum into the patient's vagina, a 5-Fr salpingographic balloon catheter (PBN Medicals, Stenløse, Denmark) was placed in the uterine cavity and filled with 1–2 mL air. This step ensured that the cervical canal was closed, thus preventing leakage of fluid and keeping the catheter in the correct position. The TVS probe was positioned to visualize the transverse section of the uterus and, if possible, both ovaries laterally. The CCI mode was then started, causing the view of the pelvis to become completely anechoic. Using CCI technology during HyCoSy, the intrauterine injection of ultrasound contrast medium in a completely anechoic pelvis was visualized as hyperechoic fluid, seen first in the uterus, then in the tube if it was proximally patent and, finally, as it spilled into the abdominal cavity if the tube was completely patent distally. 3D volume acquisition was performed during injection, with the region of interest set as wide as possible so that uterus and both ovaries could be seen and allowing the whole length of the Fallopian tube to be detected.

The sonographic volume acquisition technique was standardized according to the following criteria: frequency, 6–9 MHz; sweep angle, as wide as possible to include the uterus in transverse section and both ovaries laterally (maximum 120°); sweep velocity, medium to maximum quality; 3D box size, exceeding the uterus and both ovaries by 1 cm on each side. The contrast medium was highly echogenic and could be visualized for several minutes, allowing the tubal course and shape to be studied and more than one 3D volume acquisition to be performed. For each patient, 3D volume acquisition was performed consecutively twice, while injecting up to 4 mL contrast media: 1.5 mL SonoVue diluted with 2.5 mL saline solution the first time and 2–4 mL saline (which pushed the residual contrast agent through the catheter into the tubes) the second time.

The multiplanar view of the uterus and tubes, obtained during injection of the contrast medium (Figure 1), was converted automatically, by dedicated software, to the volume image. This resulted in a view of the uterine cavity in coronal section, with both tubes laterally and the contrast medium spilling around the ovaries if both tubes were patent. The possibility to rotate this volume showed the tubal course in 3D space (Figure 2 and Videoclip S1). In addition, the software permitted the sonographer to combine gray-scale imaging of the uterus in coronal section, with automated reconstruction of the contrast medium in the uterine cavity and patent tubes superimposed in red (Figure 3 and Videoclip S2). In cases of unilateral tubal occlusion, the tubal course was seen on one side only (Figure 4b) and was not detectable if both tubes were occluded (Figure 4c). Using 2D gray-scale ultrasound of the ovary after contrast medium injection in parallel with CCI permitted visualization of the contrast agent around the ovaries as a hyperechoic ring in anechoic tissue (Figure 5), thus confirming tubal patency and spillage of the contrast fluid in the pelvis.

Figure 1.

Automated three-dimensional coded contrast imaging hysterosalpingo-contrast sonography: multiplanar view showing contrast medium in uterus and tubes and automated reconstruction of the volume.

Figure 2.

Automated three-dimensional (3D) coded contrast imaging hysterosalpingo-contrast sonography. (a) Automated reconstructed volume of contrast medium in uterus and patent tubes with spillage of contrast around the ovaries. (b,c) The possibility to rotate this volume shows the tubal course in 3D space.

Figure 3.

Automated three-dimensional coded contrast imaging hysterosalpingo-contrast sonography: gray-scale coronal section of uterus, with automated reconstructed volume of contrast medium in uterine cavity and patent tubes superimposed in red.

Figure 4.

Automated three-dimensional coded contrast imaging hysterosalpingo-contrast sonography: automated reconstructed volumes of contrast medium in uterus and tubes in cases of bilateral tubal patency (a) and unilateral (b) and bilateral (c) tubal occlusion.

Figure 5.

Visualization of the ovary with conventional two-dimensional gray-scale transvaginal sonography (a,c) and with coded contrast imaging (CCI) mode (b,d), prior to (a,b) and following (c,d) injection of SonoVue contrast agent. Note the hyperechoic contrast around the ovary on CCI after injection (d), confirming patency of the ipsilateral tube.

To determine final tubal patency status we then performed real-time 2D HyCoSy, using at least one injection of 4–5 mL saline mixed with air and noting the microbubbles' movement throughout the tubes; if a tube was patent on 3D-CCI HyCoSy and not initially on 2D HyCoSy, which may occur due to tubal spasm, we administered a further injection or waited a few seconds and reassessed patency until the patency seen on 3D imaging was confirmed. Determination of tubal patency on 2D HyCoSy required all three of the following criteria[4, 17, 20]: passage of air and saline bubbles through the interstitial part of the tube; detection of air bubbles moving around the ovary (observation of flow around the ovaries was possible even without visualization of passage through the whole tubal course); detection of the solution and air bubbles in the pouch of Douglas.

Immediately after the procedure, all patients received antibiotic prophylaxis as a single 1-g dose of azithromycin, administered orally.

Patients were asked to grade the discomfort experienced with the aid of a 10-point visual analog scale (VAS), with 0 corresponding to ‘no pain experienced’ and 10 corresponding to ‘maximum pain experienced’. The pain score was evaluated during the procedure and 5–10 min after. The need for analgesic drugs after the procedure or symptoms of a vagal reaction (nausea, bradycardia, sweating, hypotension) during the procedure were also recorded.

To determine the feasibility of the method in the evaluation of tubal patency we therefore considered: visualization of tubal patency during first and second injections of SonoVue and 3D volume acquisition; visualization of the contrast agent around the ovaries with CCI; detection on 2D real-time HyCoSy of bubbles moving in the tube, around the ovaries and in the pouch of Douglas; pain during and after the procedure (using 0–10 VAS); and other side effects (vagal reactions, need for analgesic drugs).

Statistical analysis

Final tubal patency status was determined by considering all results, after completion of the procedure (both 3D and 2D evaluations and all injections), i.e. if passage of fluid in the tube was observed by either 3D or 2D HyCoSy, we considered the tube patent. This means that we could have only false-positive occluded tubes with no false-negative occlusions. For the purposes of analysis, all 126 patients were considered to have two tubes even if a previous salpingectomy had been performed.

Statistical analysis was conducted to assess whether tubes found to be patent or occluded at 3D-CCI HyCoSy had the same status on 2D HyCoSy. This was done by determining the concordance rate of tubal patency status at first and at second 3D volume acquisitions with that at final 2D real-time HyCoSy, by calculating accuracy (percentage agreement) and Cohen's kappa index for each. The percentage of time that the contrast medium was visualized around the ovaries was also calculated, as was the mean degree of pain reported during and after the procedure and the incidence of other side effects.

Descriptive analysis was achieved using proportions, means and SD and statistical analysis was performed using Student's t-test for mean and SD. Patients with primary infertility (defined when patients had never been pregnant) or secondary infertility (defined when patients had had at least one previous spontaneous pregnancy) were placed in separate subgroups, and proportions were compared with chi-square or Fisher's exact test, as appropriate. P < 0.05 was considered statistically significant.


From March 2009 to February 2010 we enrolled 131 patients into the study. Five patients were excluded because of cervicovaginal infection (n = 4) or hydrosalpinx (n = 1) detected at the time of preliminary TVS. Thus, 126 patients, 83 with primary and 43 with secondary infertility, underwent 3D-CCI HyCoSy followed by 2D real-time HyCoSy.

Population characteristics are summarized in Table 1. Patients with primary infertility differed significantly from those with secondary infertility in only a few studied features, with a higher prevalence of endometriosis in patients with primary infertility and, as might be expected, the occurrence of miscarriages and ectopic pregnancies only in the secondary infertility group.

Table 1. Population characteristics in 126 infertile patients who underwent hysterosalpingo-contrast sonography, according to level of infertility
CharacteristicPrimary infertility (n = 83)Secondary infertility (n = 43)PaTotal (n = 126)
  1. Data are given as mean ± SD or n (%).

  2. a

    Primary vs secondary infertility patients (Student's t-test, chi-square test or Fisher's exact test). TVS, transvaginal sonography.

Age (years)34.8 ± 4.438.4 ± 4.3<0.000136.1 ± 4.7
Body mass index (kg/m2)20.9 ± 2.922.2 ± 2.80.01721.4 ± 2.8
Presence at initial TVS of:    
  Endometriosis15 (18.1)2 (4.6)0.0517 (13.5)
  Myomas10 (12.1)6 (13.9)0.7816 (12.7)
History of:    
  Ectopic pregnancy0 (0)7 (16.3)0.00047 (5.5)
  Pelvic inflammatory disease3 (3.6)0 (0)0.553 (2.4)
  Recurrent miscarriage0 (0)9 (20.9)<0.00019 (7.1)
  Salpingectomy1 (1.2)5 (11.6)0.0176 (4.8)
  Myomectomy3 (3.6)2 (4.6)1.005 (3.9)
  Metroplasty1 (1.2)1 (2.3)1.002 (1.6)
  Surgery for endometriosis3 (3.6)1 (2.3)1.004 (3.2)

At the end of the entire procedure (both 3D and 2D real-time HyCoSy) we observed 111 patients with bilateral tubal patency, 11 patients with unilateral tubal patency and four patients with bilateral tubal occlusion, giving a total of 232 (92%) patent and 20 (8%) occluded tubes (252 tubes in 126 patients). We considered final tubal patency status results from primary and secondary infertile patients together because they were not statistically significantly different. Tubal patency status results at the first and second injections of contrast agent and 3D volume acquisition and the final result after 2D real-time HyCoSy are shown in Figure 6 and Table 2. The first and second 3D volume acquisitions in relation to final tubal patency status on 2D evaluation had accuracy (concordance rate) of 84% and 97%, respectively, and Cohen's kappa index of 27% and 68%. Bilateral tubal patency was observed in 79 (62.7%) patients at first injection and 3D volume acquisition, in 104 patients (82.5%) at second injection and 3D volume acquisition, and finally was confirmed in 111 patients (88%) at 2D real-time evaluation. In only eight patients were further injections and 2D real time examination necessary to achieve the correct final tubal patency status (Figure 6); the concordance rate to the final TVS tubal patency status for tubal status following the second injection and 3D volume acquisition was 94% (Table 2).

Figure 6.

Flow chart of tubal patency status observed in 126 patients during hysterosalpingo-contrast sonography (HyCoSy) at first and second SonoVue injections with three-dimensional coded contrast imaging (3D-CCI) volume acquisition, and after two-dimensional (2D) real-time HyCoSy final evaluation. Numbers of patients who had incorrect classification of their tubal status at the first 3D evaluation and needed further injections are given in the small dotted-line boxes.

Table 2. Tubal patency status results in 126 infertile patients (252 Fallopian tubes) after first and second SonoVue contrast medium injection with three-dimensional coded contrast imaging (3D-CCI) evaluation, and after two-dimensional (2D) real-time hysterosalpingo-contrast sonography (HyCoSy) final evaluation
Tubal patency statusHyCoSy modality
3D-CCI volume acquisition2D real-time final evaluation
  1. Data are given as n (%) of patients or tubes.

  2. a

    Concordance rate between each 3D volume acquisition and final 2D real-time evaluation.

Patients (n = 126)   
  Bilaterally patent79 (62.7)104 (82.5)111 (88.1)
  Bilaterally occluded12 (9.5)5 (4.0)4 (3.2)
  Unilaterally occluded35 (27.8)17 (13.5)11 (8.7)
  Concordance rate to final resultsa94 (74.6)119 (94.4)126 (100)
Tubes (n = 252)   
  Patent tube193 (76.6)225 (89.3)233 (92.5)
  Occluded tube59 (23.4)27 (10.7)19 (7.5)
  Concordance rate to final resultsa212 (84.1)244 (96.8)252 (100)

Bilateral tubal occlusion was confirmed by laparoscopy in all four patients with bilateral occlusion on 2D HyCoSy. Among the 11 patients with unilateral tubal occlusion we had no laparoscopic confirmation of tubal patency status; however, five of these patients had a previous unilateral salpingectomy for ectopic pregnancy or endometriosis and in one patient HSG confirmed unilateral tubal occlusion. In one case of previous salpingectomy for ectopic pregnancy we detected contrast media and fluid passage; this was in fact likely through the residual proximal tube.

Contrast agent around the ovaries was observed in 121 out of 126 patients (96%) (Figure 5). It was detected bilaterally in all cases of bilateral tubal patency except one (spillage of contrast medium in the peritoneal cavity was observed but it did not surround both ovaries clearly, possibly due to the presence of adhesions). In four of the 11 patients with unilateral tubal occlusion it was detected bilaterally, even around the ovary ipsilateral to the occluded tube (perhaps because of diffusion of fluid in the pelvis). Contrast agent spillage around the ovaries was absent in the four patients with bilateral tubal occlusion.

A pain score >5 (on 0–10 VAS) was recorded in 57.9% of patients during the procedure, but a pain score ≤ 5 was recorded in 85.7% of patients immediately after the procedure (Table 3), with only 11% of patients (n = 14) requiring analgesic drugs and only 5.6% (n = 7) with vagal reactions. All seven of the latter patients responded well to atropine administration. In no case was hospital admission required.

Table 3. Pain evaluation by visual analog scale (VAS) in 126 infertile patients during and after hysterosalpingo-contrast sonography (HyCoSy)
 Score ≤ 5 (n (%))Score >5 (n (%))VAS score (mean ± SD)
  1. VAS score ranges from 0 to 10, with 0 corresponding to ‘no pain experienced’ and 10 corresponding to ‘maximum pain experienced’.

Pain during HyCoSy53 (42.1)73 (57.9)5.31 ± 2.8
Pain after HyCoSy108 (85.7)18 (14.3)2.41 ± 2.7
Pain during and after HyCoSy55 (43.6)13 (10.3)3.86 ± 2.0
Vagal reactions2 (1.6)5 (4.0) 
Analgesic drugs administered2 (1.6)12 (9.5) 


Several studies[4-9, 17-21] attest to the high accuracy of transvaginal HyCoSy in the assessment of tubal patency compared with laparoscopic chromopertubation and HSG. However, 2D HyCoSy, even when performed with ultrasound-enhancing contrast media, is a highly operator-dependent technique. An experienced sonographer is necessary to manipulate the TVS probe efficiently in order to detect the contrast medium echoes and visualize the different parts of the tubes, as only rarely can the entire tube be visualized during fluid injection. 3D acquisition starting from a transverse section of the uterus is static, does not require probe movement and automatically generates the multiplanar image and the reconstruction of the entire uterine cavity and bilateral tubal course in case of patency.

Combination of a second-generation contrast medium with CCI, which discriminates the harmonic response of contrast microbubbles from ultrasound echoes of surrounding tissues, is very helpful for clearer visualization of the fluid passage throughout the tubes. Recently, high accuracy (90%) of automated 3D-CCI HyCoSy in detecting tubal patency has been reported[21], but the procedure seems difficult to perform for many sonographers.

With this feasibility study, therefore, we wanted to assess first how simple it is to perform automated 3D-CCI HyCoSy and second if it is as accurate as the more difficult 2D HyCoSy. Concordance with final tubal patency status of 97% was obtained at the second 3D volume acquisition, demonstrating that when tubal patency by 3D ultrasound is observed, no further evaluation is needed. With previous studies we have demonstrated that HyCoSy has a high concordance rate to laparoscopy[4], which is still considered the gold standard for evaluation of tubal patency status.

If tubal occlusion is detected during 3D ultrasound, further injections are required, especially in cases of unilateral tubal occlusion. In many of these cases, the fluid passage is better in one tube whereas the other tube seems to be occluded, possibly due to spasm. After waiting some minutes and repeating the injection, often the tube is seen to be patent. However, on repeated injections, 3D volume acquisition cannot show accurately the tubal course, because spillage of fluid in the peritoneal cavity provides a diffuse image of contrast medium around the pelvic organs and not just in the tube.

Moreover, with 2D real-time HyCoSy, for each tube one injection and often more than one is needed to visualize the passage of fluid in the tubes bilaterally. Because both tubes are visualized simultaneously during 3D HyCoSy volume acquisition with only one or two injections, less contrast agent and fluid is required in comparison to 2D real-time HyCoSy. In our experience, and in accordance with other studies[6, 22], the greater the volume of fluid injected, the more pain is felt by the patient; however, there is no association with the incidence of vagal reactions. While more than 50% of our patients reported pain during the procedure, it disappeared in most cases immediately afterwards. This was probably because the cramping and discomfort was due not only to the injections but also to the balloon catheter inside the uterine cavity.

Another major advantage of HyCoSy with automated 3D-CCI is the possibility to obtain images of the uterus and tubal course that are similar to those obtained by HSG, with the advantage that the volume can be moved and rotated in 3D space. They can also be shared easily with other clinicians, whereas 2D HyCoSy is a dynamic real-time exam in which the passage of fluid is seen only by the ultrasound examiner and images are very difficult to interpret. Moreover, the acquired volumes can be stored and analyzed later, reducing the examination time for patients.

A disadvantage of this new technique compared with 2D HyCoSy is the cost, of both a 3D machine and the contrast medium. Another limitation is the fact that most ultrasound contrast media are approved in many countries only for intravenous injection, and not for injection into the uterus and tubes.

In conclusion, HyCoSy with automated 3D-CCI technology retains the advantages of conventional 2D HyCoSy while at the same time overcoming the disadvantages. This technology permits visualization of the tubal course, creating images of the tubes in the coronal view and obtaining a volume so that the tubal course in 3D space can be evaluated. It provides imaging of the entire tube and its course, similar to HSG, with the advantages of: requiring no exposure to radiation; the possibility of repeating injections if needed; and the possibility of being performed in an outpatient setting. The clarity with which spillage is seen and the ease of automated 3D volume acquisition without any difficult probe movements makes this diagnostic method straightforward even for an inexperienced sonographer. With this study we have shown that this 3D-CCI HyCoSy method has high diagnostic feasibility, and can be performed easily in the office. It provides the possibility of infertile patients having only one scan in the proliferative phase of the cycle, while obtaining information about uterine and adnexal morphology, as well as tubal patency (One-stop Infertility Clinic). We are confident that 3D-CCI HyCoSy will find a place in the early outpatient investigation of infertile women.


The following supporting information may be found in the online version of this article:

Videoclip S1 Video showing rotation of uterine and tubal volume obtained by three-dimensional (3D) coded contrast imaging hysterosalpingo-contrast sonography and how tubal course in 3D space can be better evaluated.

Videoclip S2 Video showing automated reconstructed uterine volume in gray scale, with overlapping red contrast media in Fallopian tubes.