Technical Innovation
Women's Imaging
February 2008

A Hybrid Radiography/MRI System for Combining Hysterosalpingography and MRI in Infertility Patients: Initial Experience


OBJECTIVE. We evaluated the feasibility of a prototype hybrid radiography/MRI system in evaluating infertility patients. Pelvic MRI was followed by hysterosalpingography (HSG) without moving the patient. This system allowed evaluation of tubal patency and cross-sectional imaging with one examination.
CONCLUSION. Our hybrid radiography/MRI system provided good-quality HSG and MR images. We were able to assess tubal anatomy and patency and uterine anatomy and to detect pelvic abnormalities, including fibroids and adenomyosis. Furthermore, MR images and radiographs were superimposed to clarify HSG findings.


The imaging workup of infertility patients utilizes several techniques. Hysterosalpingography (HSG) is the gold standard for assessing fallopian tube patency [1]. However, MRI is necessary to characterize uterine anomalies and pelvic abnormalities [2]. The workup of infertility may be accomplished by several different techniques and sometimes requires multiple different examinations. A truly hybrid radiography/MRI system developed at our institution enables MRI and fluoroscopy of the same imaging volume without moving the patient [3]. This system combines HSG and MRI in one examination. Our purpose was to evaluate the clinical feasibility of obtaining images using our hybrid radiography/MRI system to evaluate infertility patients.
The truly hybrid radiography/MRI system is unique because fluoroscopy and MRI are performed without moving the patient between techniques. Other systems that combine fluoroscopy and MRI use a system whereby adjacent MRI and radiography suites are connected using a moving-bed patient system [4]. Our unit is truly hybrid in that a fixed X-ray source and digital flat-panel detector are located inside the bore of a 0.5-T magnet (Signa-SP, GE Healthcare). This fully integrated setup allows easy and rapid switching between radiography and MRI without needing to move the patient. For evaluation of infertility, HSG and pelvic MRI are performed without moving the patient, aside from elevating the legs to the lithotomy position to achieve cannulation of the cervix for HSG. The two examinations were thereby effectively combined [3], which enables evaluation of tubal patency and cross-sectional pelvic abnormalities in one examination. In addition, using the same field of view and plane of imaging allows us to superimpose HSG and MR images.
Using our hybrid radiography/MRI system, we performed pelvic MRI and HSG for the evaluation of infertility patients. We evaluated the performance and image quality and report our initial experience with this new technique.

Materials and Methods

We obtained institutional review board approval for our study and informed consent from patients, and the study was compliant with HIPAA. We recruited patients who were scheduled to undergo HSG for infertility. One patient was examined for uterine dehiscence. Examinations were performed by two radiologists: an experienced body imager and a body imaging fellow.
The hybrid radiography/MRI system is a prototype created at our institution [3]. A fixed-anode fluoroscopic X-ray tube and digital flat-panel detector are positioned between the two magnetic poles of a 0.5-T magnet (Signa-SP) as shown in Figure 1. A prototype X-ray-transparent pelvic phased-array coil was used [5]. With patients in the supine position, axial STIR (TR/TE, 4,016/61), coronal T1 (450/18), sagittal T2 fast spin-echo (3,800/105), and coronal T2 fast spin-echo (3,000/111) sequences were acquired. Patients were then placed in the lithotomy position. The cervix was cannulated using sterile technique and MR-compatible equipment. A 6.8-French 30-cm balloon HSG catheter (Sholkoff balloon HSG catheter, Cook OB/Gyn) was advanced and the balloon was inflated. The patient was returned to the supine position and a second coronal T2 fast spin-echo sequence was performed to correspond to the subsequent HSG images. Catheter positioning was confirmed with this sequence. Iohexol (Omnipaque 240, GE Healthcare) was injected under fluoroscopic guidance until bilateral free peritoneal spill was observed. Tubal patency, uterine anatomy, ovaries, and other abnormalities were assessed. When appropriate, MR and HSG images were superimposed. Image quality was assessed—in particular, with regard to whether the fluoroscopic images were adequate to be used to evaluate tubal patency.
Fig. 1 Photograph of hybrid radiography/MRI system shows position of X-ray tube (thin black arrows) and flat-panel detector (long white arrow) between two donut-shaped magnetic poles (thick black arrows) of 0.5-T magnet (Signa-SP, GE Healthcare). Foot pedals for fluoroscopy are indicated with short white arrow. Patients were placed in supine position for acquisition of MR images. For cannulation of cervix, patient was placed in lithotomy position using stirrups attached to inside of magnet.


Ten patients were examined. Nine were referred for infertility and one for uterine dehiscence after cesarean delivery. An example of a normal study with a superimposed composite of the MR and HSG images is shown in Figure 2. The fluoroscopic images were of good quality in all but one case. Inadequate X-ray tube output in an early prototype was the cause of poor image quality in that study. In the other nine cases, bilateral tubal patency was shown, with good resolution of the contrast-filled tubes and peritoneal spill. In two patients, contour abnormalities of the endometrial cavity were detected.
Thirteen fibroids in five patients were shown on MRI. In four patients, the fibroids were undetectable on HSG, including one case with bulky, exophytic lesions. In the fifth patient, an intramural fibroid abutting the endometrium distorted the endometrial cavity, causing a contour abnormality on HSG (Figs. 3A and 3B). Superimposition of the MR and fluoroscopic images confirmed that the abnormality on HSG corresponded to the fibroid (Fig. 3C).
In another case, HSG showed a fundal contour abnormality that was interpreted as an arcuate uterus. MRI showed that the contour abnormality corresponded to an area of focal adenomyosis (Figs. 4A and 4B). The contour abnormality on HSG was associated with subtle irregularity, likely representing diverticula extending into the myometrium, which is a documented HSG finding suggestive of adenomyosis [6].
Cervical defects were seen on MRI in two patients from surgical procedures. In one case, the lesion was retrospectively visible on HSG. In a third patient, a cesarean delivery defect and possible fistulous tract were noted. MRI showed malposition of the HSG catheter in three patients. After the catheter was repositioned, MR images were obtained to confirm its proper placement.
Fig. 2 Normal study in 36-year-old woman with infertility. Superimposition of coronal MR image and hysterosalpingogram (red) can be performed because two techniques have same field of view and are obtained without moving patient between studies.
Vertical shadowing artifact was noted in some fluoroscopic images. This artifact appeared because images obtained using our prototype system are not corrected for gain due to pixel-to-pixel variation. This artifact can be removed by postprocessing; however, we did not perform postprocessing because images with the artifact remained of diagnostic quality.


The most common female component of infertility is related to tubal patency [7]. Most clinical MRI techniques do not provide the spatial resolution necessary to show the fallopian tubes. Therefore, HSG remains the gold standard for the evaluation of tubal patency. MRI remains the gold standard for evaluating uterine contour abnormalities [2]. The low specificity of HSG for contour abnormalities—in particular, for distinguishing bicornuate from septate uteri—is well known [8]. Our data support the results of previous studies that show MRI is more sensitive than HSG for identifying fibroids. Currently, therefore, both techniques are often necessary.
We believe that our hybrid radiography/MRI system offers the advantage of incorporating two techniques into one system, thereby streamlining the workup of infertility and identifying unsuspected abnormalities. An exciting opportunity provided by the hybrid radiography/MRI system is the exact coregistration of the two techniques. As we have shown, this system has the potential to show the cause of hysterographic contour abnormalities with cross-sectional images of myometrial abnormalities, such as fibroids and adenomyosis. Improved image quality and coregistration of images may increase the utility of this new technology.
Fig. 3A 40-year-old woman with infertility. Hysterosalpingography (HSG) image shows subtle contour abnormality (arrows) in lower uterine segment.
Fig. 3B 40-year-old woman with infertility. HSG finding shown in A corresponds to intramural fibroid (arrow) that abuts endometrial lining on coronal T2 fast spin-echo MR image. MR image clearly shows that partially submucosal fibroid is responsible for contour abnormality shown on hysterosalpingogram (A).
Fig. 3C 40-year-old woman with infertility. Superimposition of HSG (red) and MR images shows coregistration of contour abnormality on HSG and fibroid on MRI.
Fig. 4A 42-year-old woman with infertility. Focal contour abnormality in uterine fundus (arrows) seen on hysterosalpingography (HSG) image was interpreted to be arcuate uterus.
Fig. 4B 42-year-old woman with infertility. Coronal T2 fast spin-echo MR image confirms that contour abnormality (arrow) shown on HSG (A) corresponds to area of high T2 signal intensity, which is consistent with focal adenomyosis.
A limitation of this study is small sample size. This report is intended as a description of our initial clinical experience. We intend to increase our experience and sample size. Another limitation is that none of our patients had a significant uterine anomaly. Such a finding would emphasize the advantage of incorporating MRI into the initial imaging workup. Several technical limitations are inherent to this study: The use of a 0.5-T magnet results in decreased signal-to-noise ratio, lower spatial resolution, decreased imaging speed, and inability to perform sequences with conventional chemical fat suppression when compared with 1.5-T MRI. The limitations of our radiography unit are the fixed projection angle and lower power compared with rotating tubes. Although this poses a limitation to angiographic studies, the power is sufficient for HSG.
In conclusion, our initial experience shows that our unique, truly hybrid radiography/MRI system provides information regarding tubal patency, uterine anatomy, and pelvic abnormalities in one examination. Furthermore, MR images and radiographs can be superimposed to clarify HSG findings. Our study confirmed that uterine abnormalities, such as fibroids and adenomyosis, can be missed on HSG. Based on this initial experience, we believe that our hybrid radiography/MRI system has the potential to streamline the imaging workup of infertility patients. Additional studies are necessary to determine the incremental benefit of the new technique compared with conventional hysterography in identifying unsuspected abnormalities.


Address correspondence to C. B. Freeman-Walsh.
Supported by National Institutes of Health grant R01 EB000198.
This is a Web exclusive article.


Yoder IC, Hall DA. Hysterosalpingography in the 1990s. AJR 1991; 157:675 –683
Imaoka I, Wada A, Matsuo M, Yoshida M, Kitagaki H, Sugimura K. MR imaging of disorders associated with female infertility: use in diagnosis, treatment, and management. RadioGraphics 2003; 23:1401 –1421
Ganguly A, Wen Z, Daniel BL, et al. Truly hybrid X-ray/MR imaging: toward a streamlined clinical system. Acad Radiol 2005; 12:1167 –1177
[No authors listed]. Philips Medical Systems builds world's first XMR system at UCSF. JBR–BTR 2001; 84:302
Rieke V, Ganguly A, Daniel BL, et al. X-ray compatible radiofrequency coil for magnetic resonance imaging. Magn Reson Med 2005; 53:1409 –1414
Simpson WL, Beitia LG, Mester J. Hysterosalpingography: a reemerging study. RadioGraphics 2006; 26:419–431
Hornstein MD, Schust D. Infertility. In: Berek JS, Adashi EY, Hillard PA, eds. Novak's gynecology, 12th ed. Baltimore, MD: Williams & Wilkins, 1996:915 –962
Dudiak CM, Turner DA, Patel SK, Archie JT, Silver B, Norusis M. Uterine leiomyomas in the infertile patient: preoperative localization with MR imaging versus US and hysterosalpingography. Radiology 1988; 167:627 –630

Information & Authors


Published In

American Journal of Roentgenology
Pages: W157 - W160
PubMed: 18212200


Submitted: March 20, 2007
Accepted: September 13, 2007


  1. gynecology
  2. hysterosalpingography
  3. infertility
  4. MRI
  5. obstetrics
  6. women's imaging



Cynthia B. Freeman-Walsh
Department of Radiology, Stanford Hospital, Stanford, CA.
Present address: Department of Radiology, The Ottawa Hospital, 1053 Carling Ave., Ottawa, ON K1Y 4E9, Canada.
Rebecca Fahrig
Department of Radiology, Lucas MRS Center, Stanford, CA.
Arundhuti Ganguly
Department of Radiology, Lucas MRS Center, Stanford, CA.
Viola Rieke
Department of Radiology, Lucas MRS Center, Stanford, CA.
Bruce L. Daniel
Department of Radiology, Stanford Hospital, Stanford, CA.

Metrics & Citations



Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options


View PDF

PDF Download

Download PDF







Copy the content Link

Share on social media