Reproductive Endocrinology Diagnostic Imaging

Peter Klatsky and Victor Y. Fujimoto Introduction

The last few decades have witnessed rapid advances in diagnostic and therapeutic options in modern medicine. While the field of reproductive endocrinology has stayed away from many diagnostic modalities using ionizing radiation, the field has benefited from advances in imaging modalities such as ultrasound and magnetic resonance. In this chapter, we will review the general principles of these techniques and offer examples from reproductive endocrinology where imaging has proved helpful.

Principles of Ultrasound and Magnetic Resonance Imaging (MRI)

Sonographic imaging relies on the different transmitting and reflecting properties of ultrasound waves through various types of tissues. Ultrasound waves are transmitted and reflected waveforms received through the transducer, ultimately leading to the creation of images.

An important setting on an ultrasound transducer is the frequency of the ultrasound waves transmitted. Lower frequency sound waves are better able to transmit through deeper tissue but have less clarity. These frequencies are often employed in transabdominal imaging and are particularly helpful in larger patients with deep adipose tissue. Abdominal ultrasound usually utilizes 3-5 MHz in transmitting images. Scans that require less penetration but greater detail, such as transvaginal ultrasound, are best obtained by using higher frequency sound waves, usually 7 MHz.

Abdominal ultrasound is best performed in patients with a full bladder, where the bladder functions as a window through which ultrasound waves can travel unimpeded to penetrate underlying pelvic structures. Transvaginal ultrasound functions best in patients with empty bladders so that the uterus lies in its naturally-flexed position and is not artificially displaced away from the vagina. An anteverted uterus is more likely to place the ovaries closer to the transvaginal transducer.

Magnetic resonance takes advantage of the physical principle that the nuclear spin of hydrogen ions varies depending on its chemical environment. MRI utilizes radiofrequency to measure this spin and its variances throughout the body to create excellent images of soft tissue structures and planes between tissues.

Ambiguous Genitalia

Diagnostic imaging assists providers in making a rapid diagnosis and planning appropriate gender assignment and treatment strategies. The finding of ambiguous genitalia is considered by the American Academy of Pediatrics to be a pediatric emergency. Electrolyte abnormalities from severe congenital adrenal hyperplasia can be life-threatening. Similarly urgent is the social crisis that results in long-term stig mata that develop when parents are unable to tell friends and family whether they have just had a baby boy or girl.

The initial workup to determine the infant's gender includes blood tests to assess hormone levels, electrolyte concentrations, as well as karyotyping. Unfortunately, the results of some these tests, (e.g., karyotyping) can take 48 hours or more to obtain results.

An abdominal ultrasound can provide immediate data to assist a family in understanding their newborn's phenotype, likely gender assignment and reproductive potential. An abdominal ultrasound should be immediately performed to assess the presence of a cervix, uterus, fallopian tubes, and ovaries or gonads. Evidence of a uterus on ultrasound is the most important finding and will most likely reassure parents that their newborn will develop into a phenotypic female. The most common cause of virilization of a female infant at birth is congenital adrenal hyperpla-sia, which can be managed medically by replacing cortisol.

Other causes of ambiguous genitalia include gestational hyperandrogenism often related to maternal hyperandrogenism during pregnancy. A genotypic male infant with 5-alpha reductase deficiency can also present initially as a minimally-virilized female infant.

While ultrasound is the first step, nondiagnostic exams are not uncommon, and MRI can also provide useful information. A diverse array of mullerian anomalies can be identified by MRI with its improved ability to identify the cervix, uterus, and gonads. Many of these structures cannot be seen consistently on ultrasound. One study demonstrated that MR could identify the uterus in 93% of patients, the vagina in 95%, the penis in 100%, the testis in 88%, and the ovary in 74% of patients.

The advantage of MRI is that it can better elucidate soft tissue structures through use ofT1- and T2-weighted sequences and its ability to look with equal clarity through multiple depths of tissue. MR also offers larger visual fields than ultrasound, with multiple planes of images that can be viewed simultaneously to identify and correlate related structures. In addition to assisting with rapid determination of accurate gender assignment, MR is also useful in planning surgical reconstruction, such as with a transverse vaginal septum or aplastic vagina in Mayer-Rokitansky-Kuster-Hauser (MRKH) syndrome.


Many cases of mullerian anomalies present for the first time during adolescence with primary amenorrhea (Figs. 12.1-12.5). Fifteen percent of these women will have abnormal pelvic exams. Breast development, normal growth and pubarche in the presence of a blind-ending or absent vagina suggests either mullerian agenesis, transverse vaginal septum, or imperforate hymen. Lagging thelarche or pubarche in a young woman with tall stature and a blind-ending vagina may suggest androgen insensitivity syndrome.

After a physical exam, ultrasound and karyotyping are the first steps to assess for a uterus, vagina, or gonads. Any patient with a female phenotype and 46 XY genotype must have her gonads removed after completion of puberty or sooner depending on the condition as they are at risk for malignancy.

Distinguishing a transverse vaginal septum from an imperforate hymen is important in determining an appropriate and safe therapeutic approach. While ultrasound is helpful, MRI is the gold standard for identifying corresponding structures and planning surgical treatment (Fig. 12.3).

Rokitansky Syndrome
Figure 12.1. MRI showing sagittal view of a patient with Mullerian agenesis or Mayer-Rokitansky-Kuster-Hauser syndrome. Note the absence of a uterus posterior to bladder.

Mullerian agenesis (Mayer-Rokitansky-Kuster-Hauser syndrome) results from the abnormal or absent development of the mullerian duct structures (Fig. 12.1). Patients with this condition display a variety of anomalies, most commonly complete or partial agenesis of the uterus, vagina, or cervix. About 10% of patients have a uterus but an incomplete outflow track and therefore present with cyclic pain but absent vaginal bleeding (Figs. 12.2-12.4). Ovarian function is preserved in all of these patients and can be evidenced by normal pubertal growth and development, as well as by measuring basal body temperatures and serum progesterone levels to assess ovulation.

Any patient with mullerian anomalies needs an abdominal ultrasound, MRI, or intravenous pyelogram to examine the kidneys and ureteral development (Fig. 12.5). Unilateral renal agenesis has been documented in 30% of cases, most commonly in association with a unicornuate uterus with a mullerian ductal remnant. The affected or absent kidney is generally ipsilateral to the aplastic mullerian horn.

When physical exam and ultrasound are inadequate or incompletely diagnostic, MR imaging is usually successful in characterizing mullerian anomalies and planning potential surgical treatments. In a review of 29 patients at University of

11:53:18 am 6Cg-S 16Hi 160mm



Figure 12.2. Abdominal ultrasound demonstrating hematometria associated with vaginal agenesis.

Figure 12.2. Abdominal ultrasound demonstrating hematometria associated with vaginal agenesis.

Transverse Vaginal Septum
Figure 12.3. MRI showing sagittal view of hematocolpometria in patient with transverse vaginal septum.
Hymen Imperforate
Figure 12.4. Coronal MRI of pelvis in a patient with an obstructed Mullerian duct remnant associated with uterine didelphis.

California - San Francisco, MRI correctly identified anatomic anomalies in all of the patients as confirmed by surgical or further invasive findings. The MRI results influenced or changed clinical treatment decisions in a quarter of patients who had previously undergone ultrasound imaging. Of particular value was the improved ability to identify gonads and establish that suspected adnexal lesions were actually components of obstructed mullerian duct anomalies.

MRI can also identify vaginal tissue and determine partial or complete agenesis, as well as duplication of vaginal and cervical development. Thin sections using transverse images are most helpful in delineation of these structures, and pretreatment with mild estrogen stimulation will assist in creating contrast along the vaginal wall.

Many mullerian anomalies present later in life, not with primary amenorrhea, but in the workup of primary infertility or recurrent pregnancy loss. MRI may also be useful in the evaluation of these women, for its ability to identify and distinguish the fibrous tissue of a septate uterus from the myometrium of bicornuate uterus. The former patients can benefit from hysteroscopic resection of their septum; the latter patients cannot. It is important to distinguish an arcuate or didelphic uterus from a septate uterus before recommending treatment. A uterine septum can be safely resected, decreasing the risk of recurrent miscarriage. An arcuate uterus is a normal variant, without negative impact on reproductive outcome and does not require additional abdominal imaging. Conversely, patients with a didelphic uterus

Arcuate Uterus And Ivf

Figure 12.5. Intravenous pyelogram showing an absent right renal collecting system as might be seen associated with mullerian anomalies.

Figure 12.5. Intravenous pyelogram showing an absent right renal collecting system as might be seen associated with mullerian anomalies.

require imaging to rule out renal anomalies. Distinguishing a fibrous septa from a patient with a didelphic uterus is important as attempting hysteroscopic resection in a patient with a didelphic or bicornuate uterus can lead to unintentional surgical perforation.

Secondary Amenorrhea

The most common cause of secondary amenorrhea in a young woman is pregnancy. If a patient has a positive pregnancy test, ultrasound can be performed to date the pregnancy and to rule out an ectopic pregnancy (the latter would be of particular concern in any patient with pain, vaginal spotting, or risk factors such as prior tubal surgery, sexually transmitted infections or pelvic inflammatory disease). A gestational sac should be visible by transvaginal ultrasound (TVUS) in most pregnancies when the serum level of human chorionic gonadotropin (phCG) is above 1500 IU/L. An empty endometrial cavity in a patient with a phCG level above 1500 IU/L might indicate an ectopic pregnancy, although a multiple gestation could not be ruled out.

In patients with secondary amenorrhea or oligomenorrhea, anovulatory conditions must be considered. These will be discussed along with appropriate diagnostic procedures in the following pages.


The workup of female infertility usually begins with an assessment of sexual and menstrual history, followed by a diagnostic workup to identify and treat a particular cause of infertility. Imaging techniques play a crucial role in the diagnosis of anatomic infertility, especially as these techniques determine the need for surgery. In recent years, imaging techniques have emerged as important in diagnosing ovula-tory defects, as well.

Anatomic and Tubal Factor Infertility

Two causative tubal factors that lead to infertility are bilateral tubal occlusion and the presence of a hydrosalpinx. Clinical risk factors for tubal factor infertility include a history of sexually-transmitted infection, pelvic inflammatory disease, tubal surgery, or severe abdominal-pelvic adhesions. The resulting adhesions block sperm or embryo transport to prevent fertilization and implantation. Hysterosalpingogram or sonohysterogram have become the first line tools for evaluation and diagnosis of tubal and anatomic causes of infertility.


One of the oldest and most widely used diagnostic imaging modalities in infertility is the hysterosalpingogram (HSG); see Figures 12.6-12.8. A hysterosalpingogram consists of a series of plain films utilizing a liquid contrast that is injected into the cervical canal. Images are taken revealing the contour of the endometrial cavity and fallopian tubes with spillage of contrast into the abdominal cavity (Table 12.1).

Hysterosalpingogram Contrast Into Cavity
Figure 12.6. Normal appearing HSG.
Hydrosalpinges Mri
Figure 12.7. HSG with unilateral tubal occlusion.
What Sonohysterogram Catheter
Figure 12.B. HSG with bilateral hydrosalpinges.

Table 12.1. Performing an HSG

1. Study should be performed on day 5-9 of menstrual cycle.

2. Premedicate with 600 mg ibuprofen (antibiotics for high risk patients).

3. Cleanse the cervix with antiseptic solution.

4. Insert HUMI or acorn tip catheter; remove open angle speculum.

5. Under fluoroscopic observation, SLOWLY instill 10 ml warm oil based contrast.

6. Take images as contrast fills uterine cavity and as soon as bilateral spill is identified.

7. May consider oblique planes for selected images.

8. Instilling contrast and taking images should take less than one minute.

9. After uterine and tubal assessment is completed, have patient roll 360° and take final image to look for fluid collections representing peritubal adhesions.

A normal HSG will demonstrate a smoothly-contoured endometrial cavity without filling defects. It will also demonstrate thin, nearly imperceptible tubes with contrast fluid spilling bilaterally. Tubal disease is suggested by absent spillage or evidence of a hydrosalpinx. With time, distal tubal occlusions can lead to the development of a hydrosalpinx which is characterized by severely damaged tubal epithelium. Both tubal occlusion and dilation can cause infertility.

Even when one tube is patent, the presence of a contralateral hydrosalpinx will decrease pregnancy rates. Presumably, the hydrosalpinx fluid has an embryotoxic effect. Infertile patients with a hydrosalpinx should be treated with salpingostomy, salpingectomy or proximal tubal occlusion. The latter two methods have been shown to improve pregnancy rates and the preferred approach has yet to be determined by a prospective, randomized trial.

Hysterosalpingography has a sensitivity of 65% and a specificity of 83% for detecting tubal disease. A significant false positive rate with the technique is partly attributed to tubal spasm during injection of contrast. Measurement of Chlamydia trachomatis antibodies has been useful in identifying additional patients at risk for tubal disease despite normal HSGs. Some providers now recommend laparoscopy with chromopertubation despite a normal HSG, for any patient with high antibody titers to Chlamydia trachomatis because of the relatively low sensitivity of HSG. Peritubal adhesions secondary to prior infection or endometriosis are often missed on HSG but can be appreciated during laparoscopy. Other authors have suggested eliminating the HSG and using only Chlamydia antibody titers to screen for tubal disease, but the low specificity and positive predictive value of titers alone continues to make HSG a clinically important study.

Although water soluble contrast agents offer somewhat clearer images, the difference in diagnostic quality is not significant. Oil-based contrasts, such as ethiodized poppy-seed oil or ethiodiol, have been associated with less pain and bleeding, and have shown a trend toward decreased infection rates.

An additional advantage to performing an HSG with oil based contrasts is the potential therapeutic benefit. Increased pregnancy rates have been documented within the first several months following the procedure. These results have not been appreciated using water soluble contrasts and for that additional reason we prefer oil-soluble media, particularly in patient populations that will not be able to afford future cycles

of in vitro fertilization. The mechanism for improved pregnancy rates after HSG with oil-based contrast is unknown, but one proposed mechanism is impairment of phagocytosis by macrophages that are present in the peritoneum and fallopian tubes and presumably interfere with fertilization. In vitro studies have repeatedly demonstrated that exposure to oil-based media impairs macrophage ability to phagocytose sperm, which may increase the chance of fertilization. Anecdotally, we have found that patients tolerate the procedure much better when warmed contrast is used and it is injected slowly. Patients should receive 600 mg of ibuprofen one hour prior to the procedure to reduce cramping and discomfort. We also avoid placing a Foley or other balloon type catheter through the internal cervical os as this can obscure intra-cavitary distortions in the lower uterine segment and increase the risk of ascending infection. An acorn tip or HUMI cannula does not share these properties. Antibiotic prophylaxis is unnecessary, although patients at high risk for infection could be given 24 hours of doxycycline after the procedure.

Abnormalities of the uterine cavity, specifically the presence of polyps, submu-cosal myomas, and fibrous uterine septa, have also been found to decrease live birth rates. Therapeutic options for intracavitary lesions include D&C, polypectomy, and hysteroscopic resection. Prior to the development of new sonographic techniques, these lesions were diagnosed by either HSG or hysteroscopy. Abnormalities on HSG appear as space occupying lesions distorting the endometrial cavity.


In many practices, sonohysterograms (SoHGs) have replaced hysterosalpingograms for screening to assess for lesions in the endometrial canal and rule out hydrosalp-inges for infertile patients considering intrauterine inseminations (IUI, Table 12.2) or in vitro fertilization (IVF, Table 12.3). SoHGs are office procedures that involve canulating the cervix with a small catheter and instilling approximately 10 ml of saline (Table 12.4). This procedure allows real time imaging and is quicker and less painful than an HSG. Figures 12.9-12.17 demonstrate the types of clinical information that can be obtained by SoHG compared to other imaging techniques.

Table 12.2. Pretreatment imaging for IUI

1. HSG and Chlamydia trachomatis titers.

2. Laparoscopy with chromotubation if elevated Chlamydia titers or endometriosis (especially with "kissing ovaries" sign).

3. If endometrial abdominally, consider SoHG.

4. If large, obscuring fibroids, consider MRI.

Table 12.3. Pretreatment imaging for IVF

2. SoHG: rule out intracavitary myomas, polyps, and hydrosalpinx.

3. Antral follicle count.

4. Assessment of endometrial thickness and morphology for embryo transfer.

Table 12.4. Performing a SoHG

1. Perform study within 10 days of LMP.

2. Patient should void prior to study.

3. Consider antibiotic prophylaxis in high risk patients.

4. Clean cervix with antiseptic solution.

5. Identify 5 Fr catheter with a 2 ml balloon.

6. Balloon catheter is prepared, tested, and primed with saline.

7. Eliminate bubbles in tubing which can create artifact on ultrasound.

8. Insert catheter; slowly inflate balloon with water.

9. Remove speculum, insert vaginal ultrasound probe.

10. Slowly instill 5-10 ml of warm saline from 20-30 ml syringe.

11. Multiple images of endometrial cavity are taken in sagittal and coronal planes.

12. Rule out hydrosalpinx.

13. Look for spill/free fluid in bilateral adnexae.

*Note: Taking care to instill fluid slowly into both the catheter balloon and the endometrial cavity will minimize patient discomfort.

A SoHG is ideally performed during the proliferative phase of the menstrual cycle when the endometrial lining is thinner in order to decrease false positive findings. Use of prophylactic antibiotics is not routinely recommended but could be considered in patients at high risk for cervicitis or subsequent PID.

Transverse Septum Uterus
Figure 12.9. Septate uterus on SoHG transverse view.
Symbols Transvaginal Ultrasound
Figure 12.11. SoHG image of endometrial polyps.


X m && ¿jJy «fr-wr*.. -ii. cSm -ft -


x ■

85dB 51/+1/Z/2 Gairi= IdB A=3


\ Jc

LMP 7/6/02


Delete Word Home Set


Figure 12.12. TVUS of endometrial polyps.

Chemest Infertile Medicine

AUG.20.04 0M11 36:42 TV G31 FR31


3=2SmN/£Znim I =?Bmm/lBwm 2= 1 Smrfi' 1 Brum 3- nms mm








AUG.20.04 0M11 36:42 TV G31 FR31


3=2SmN/£Znim I =?Bmm/lBwm 2= 1 Smrfi' 1 Brum 3- nms mm








Figure 12.13. SoHG of a submucosal fibroid.

Septate Uterus And Infertility
Figure 12.14. Axial MRI of bicornuate uterus. This may appear similar to both a septum or arcuate uterus on HSG.
Hsg Patient Images
Figure 12.15. HSG of a patient with an arcuate uterus.
Figure 12.17. Normal SoHG, transverse view.

While ultrasound is not effective at identifying tubal patency, it can accurately and comfortably assess intracavitary pathology as well as identify the presence of a hydrosalpinx. Some authors have attempted to use sonohysterograms to identify patients at risk for nondilated tubal occlusion as well.

By instilling more than 10 ml of saline and watching for accumulation of free fluid surrounding the ovary, this approach attempts to demonstrate tubal patency. Sensitivity for this test was improved by using lactose particles to increase visualization and echogenicity in spilled fluid. Unfortunately, results are unreliable and have a positive predictive value of only 40%.


Although some have attempted to use 3-D dynamic MR-hysterosalpingography to visualize both the endometrial cavity and tubes with decreased exposure to ionizing radiation from HSG imaging, we do not feel that the high costs and technical difficulties are not outweighed by any potential benefits of this modality. This is an expensive study that does not improve on sensitivity, specificity or patient comfort.

Recurrent Pregnancy Loss

Some women have no difficulties conceiving pregnancy but are plagued by recurrent early miscarriages. Common causes include chromosomal anomalies, infectious and connective tissue diseases, but the most readily treatable explanation is a problem with the uterine cavity. Anomalies of the mullerian system have been found in approximately 15% of women with three or more early pregnancy losses. The most common uterine anomaly in this group is a septate uterus. Endometrial polyps, adhesions, and submucosal myomata are also implicated in infertility. See Figures 12.9-12.14, and Figure 12.18.

It is especially important to rule out and correct any anatomic anomalies that could cause a spontaneous abortion, before undergoing in vitro fertilization. While HSG and hysteroscopy used to be the gold standard for endometrial evaluation, sonohysterography has proved more than adequate in evaluation of the uterine cavity. SoHG has a greater than 90% sensitivity and a positive predictive value of 65-100% for detecting intracavitary lesions when compared to hysteroscopy. SoHG can also be useful in detecting adhesions, or uterine synechia, that may have formed after an intrauterine procedure such as dilation and curettage (Fig. 12.18).

Unlike hysteroscopy, SoHG can reliably distinguish between an arcuate, a bicor-nuate, and a septate uterus. Sonohysterograms can also identify hydrosalpinges and evidence of adnexal masses, all of which should be excised prior to proceeding with any form of assisted reproductive technology or intrauterine insemination.

Transvaginal ultrasound without saline contrast can also be used to assess uterine morphology, but with decreased sensitivity and specificity. Endometrial polyps appear as focal areas of increased echogenicity within a wider endometrial stripe (Fig. 12.12). Unfortunately, sensitivity for detecting polyps decreases from 93.1% to 64.5% and the specificity decreases from 93.9% to 75.5% on transvaginal ultrasound without saline contrast. A divergent endometrial stripe can be evidence of a uterine septum or arcuate uterus.

Benefits of SoHG over transvaginal ultrasound include the ability to reliably assess the endometrial cavity in a patient with a myomatous uterus, in which heavy "shadowing" artifact can obscure visualization of a contiguous endometrial stripe. SoHG also better distinguishes between a hyperplastic endometrium and

Figure 12.18. Uterine cavity synechiae (adhesions) and polyps on SoHG.

an endometrial polyp, as well as identify the difference between an arcuate and a septate uterus. Similarly, HSG is unreliable at distinguishing between any of these conditions and has proven particularly unreliable at distinguishing between mullerian anomalies, such as a septum and an arcuate uterus. SoHG is twice as accurate as transvaginal ultrasound or HSG in identifying uterine cavity defects.

Ovulatory Disorders

The lack of regular ovulatory cycles, called anovulation, is another cause of infertility. The workup for anovulation is largely hormonal, but several imaging modalities can be useful as well.

Hyperprolactinemia is a treatable cause of infertility that is usually caused by a prolactin secreting pituitary adenoma or microadenoma. Since prolactin levels correlate poorly with size of an adenoma, MRI of the pituitary is indicated in any patient with an elevated prolactin. Other authors have suggested that MRI is unnecessary as the natural history of a microadenoma is unpredictable and the majority resolve or remain stable without surgery. If resources for a pituitary MRI are scarce, it is reasonable to treat empirically with bromocriptine unless the patient has signs of a macroadenoma, such as decreased visual fields (bitemporal hemianopsia) or severe headache. In that case an MRI should be sought, but a lateral film of the skull is usually sufficient to diagnose empty sella syndrome.

Polycystic ovary syndrome (PCOS) is another anovulatory condition that contributes to infertility. An international consensus conference in 2003 created a working definition for PCOS, the "Rotterdam criteria." These criteria defined PCOS as the presence of any two of three findings:

1. A history of oligo- or anovulation

2. Clinical or biochemical evidence of hyperandrogenism

Pcos Ring Pearls

Figure 12.19. Polycystic ovary with classic "ring of pearls" appearance on transvaginal ultrasound.

Figure 12.19. Polycystic ovary with classic "ring of pearls" appearance on transvaginal ultrasound.

3. Polycystic ovaries on ultrasound: Defined as 12 or more follicles measuring 2-9 mm in each ovary or ovarian volumes of greater than 10 ml.

Ovaries from a woman with polycystic ovarian syndrome classically demonstrate an abundance of peripherally situated follicles that create the "ring of pearls" appearance that typifies this syndrome (Fig. 12.19). However, authors have debated the importance of finding multiple follicular cysts on ultrasound as many fertile women with regular menstrual cycles have a similar ovarian appearance on ultrasound.

Diminished Ovarian Reserve

Diminished ovarian reserve is a term used to identify women who are nearing the end of their naturally reproductive years. In older women, this may be the result of approaching climacteric and a natural decline in follicular development. Younger women may experience a similar phenomenon called premature ovarian failure (POF) when it occurs before age 41. POF is often the result of an autoimmune attack on ovarian tissue and is associated with other autoimmune diseases such as autoimmune hypothyroidism. Other causes include exposure to ionizing radiation or alky-lating chemotherapeutic agents during treatment for cancer earlier in a woman's reproductive lives. These exposures damage developing ovarian follicles and supportive stroma, leading to POF and infertility. Women with no prior history of environmental exposures should be screened for subclinical hypothyroidism as this is the most common associated autoimmune disease.

Diagnosis of diminished ovarian reserve has traditionally been made using biochemical testing including a day 3 FSH and estradiol level, or a day 10 FSH after a clomiphene citrate challenge of 100 mg from days 5-9 of the cycle. Both of these

Normal Follicular Ultrasound Chart

Figure 12.20. Transvaginal ultrasound of a normal follicular ovary with 9 follicles.

markers were used to predict success rates for women undergoing various infertility treatments. A problem with all attempts to assess ovarian reserve is that they are nonspecific and can change with each cycle. Age alone is a good predictor and has been associated with sonographically observed decreases in mean ovarian volume and number of follicles. Most markers for ovarian reserve are used to plan expectations and outline treatment options for women with clinical infertility.

Ultrasound measurement of the ovarian antral follicle count (AFC) is a reproducible test with low inter-observer variability at clinically important levels (low AFC) and can predict response to gonadotropin stimulation. When compared with age and other biochemical markers, the AFC appears to be the single best predictor of response to IVF treatment. Patients with a basal AFC of 4 or fewer follicles have significantly higher rates of cycle cancellation (41% vs 6.4%) and lower pregnancy rates (24% vs 58%) in IVF cycles. These low pregnancy rates are primarily the result of increased cancellation rates and decreased numbers of eggs retrieved,. Furthermore, an AFC of less than or equal to two follicles and ovarian volume of less than 4 cm3 are predictive of menopausal status.

AFC is reliable in women over 35 (Fig. 12.20). Younger women have greater intercycle variation in AFC and less reliable correlation between the AFC and response to IVF. In women over 35, accurate measurement of antral follicle count can help predict responsiveness before proceeding with expensive assisted reproductive technologies. Patients with discouraging results should be counseled sympathetically about alternative methods of family development such as ovum donor cycles and adoption.

Figure 12.21. Transvaginal ultrasound of endometrioma with homogenous low level echoes.


Endometriosis is characterized by ectopic endometrial tissue outside the uterine cavity which can cause pain, adhesions and infertility. Endometriosis often presents clinically with a long history of cyclic pelvic pain that may be exacerbated by vaginal penetration.

The evidence that endometriosis is implicated in infertility derived from observations that endometriosis is present at laparoscopy in 21% of women being evaluated for infertility and only 6% of fertile women undergoing laparoscopic tubal ligation. It is estimated that 30% to 50% of women with endometriosis are infertile.

Laparoscopy remains the gold standard for diagnosis and treatment for this disorder; however as less invasive diagnostic modalities have also gained favor, noninvasive diagnostic imaging techniques have proved worthwhile as well.

The most commonly used imaging technique in endometriosis is ultrasound. Transvaginal ultrasound is useful in identifying endometrial cysts or endometrio-mas. These lesions are often found during a pelvic pain or infertility evaluation, but they can be found incidentally during imaging of the abdomen and pelvis for other indications.

Endometriomas typically appear as homogenous, hypoechoic ovarian cysts with low level echoes (Fig. 12.21). They often have septations and can occur in multiples or as a single cystic mass. While mural nodularity suggests neoplasm, these nodules must be distinguished from hyperechoic wall foci, a finding which correlates strongly

Figure 12.22. "Kissing ovaries" sign. Note how dense adhesions bring the ovaries together in the posterior cul de sac.

Figure 12.22. "Kissing ovaries" sign. Note how dense adhesions bring the ovaries together in the posterior cul de sac.

with endometriomas. These densities are usually smaller and more echogenic than those associated with neoplasia. In the presence of septations, low level internal echoes, and hyperechoic wall foci, a multiloculated mass is 64 times more likely to be an endometrioma than any other adnexal mass.

With in vitro fertilization, simple endometriomas need not be removed prior to proceeding with a cycle. Care should be taken to avoid entering the endometrioma during egg retrieval as its contents are potentially toxic to the ovaries.

The "kissing ovaries" sign is an interesting sonographic feature seen in patients with severe endometriosis and pelvic adhesions (Fig. 12.22). Identifying adjacent and "kissing" ovaries at ultrasound is a strong marker for the presence ofsevere endometriosis. This is one of the few times that ultrasound can be reliably used to diagnose pelvic adhesions. One study demonstrated a positive predictive value of greater than 90% for endometrial implants involving the bowel and fallopian tubes in these patients. Other signs of adhesions include posteriorly-displaced uterus and ovaries which may be appreciated by ultrasound as a fixed, retroverted uterus. Thus, these findings can eliminate the need for further invasive diagnostic procedures, and laparoscopy can be avoided in a patient wishing to attempt hormonal control of her disease.

A concerning characteristic of endometriomas is the presence of hyperechoic deposits in the cyst wall. Nevertheless it can be difficult to distinguish between true endometriomas and neoplasms. In order to better characterize an adnexal mass that may be a neoplasm or an endometrioma, both CT and MRI can be used. When available, MR imaging avoids exposure to ionizing radiation and can improve the diagnostic accuracy and specificity of ultrasound. MR images are not as badly affected by adhesive disease and allow for a larger field of view than ultrasound.

Magnetic resonance imaging may reduce the need for invasive surgical diagnostic procedures. Endometriosis appears as a region of a low signal intensity on T1 within an area of high signal intensity on T2. High intensity nodules on the bowel or bladder suggest involvement of these tissues. An additional advantage of MR over laparoscopy is its ability to show extraperitoneal sites of involvement and lesions that would be obscured by dense adhesions on laparoscopy. Unfortunately, MR performs poorly at diagnosing peritoneal adhesions that do not involve the ovaries. MRI and laparoscopy can therefore be complementary in severe cases.


Leiomyomata are benign smooth muscle neoplasms. The predominant symptoms in patients with large fibroids are menorrhagia and pelvic pressure. They are usually diagnosed by ultrasound or physical exam, but ultrasound can be limited in a very enlarged uterus with multiple myomas. In an enlarged uterus, MR can help identify clinically important submucosal myomas and plan surgical treatment.

SoHGs are helpful in identifying location of fibroids and whether they can be removed hysteroscopically. Hysteroscopic removal is attempted in all patients with submucosal fibroids with at least half of their volume in the endometrial canal.

The clinical significance of intramural and subserosal fibroids is controversial. Some authors suggest myomectomy for patients with repetitive failed IVF cycles and no explanatory factors except intramural fibroids. Large fibroids that distort the endometrial canal are particularly worrisome. Ultrasound and SoHG have been recommended to evaluate the endometrial cavity and location of women with small to moderate-sized fibroids. Women with large fibroids can benefit from MRI studies in order to assist in preoperative visualization and surgical planning. MR is also useful in identifying the endometrial stripe and ovaries in patients whose large myo-mas cause too much shadowing and artifact to accurately assess the endometrium or adnexae (Fig. 12.23).

Although extremely rare, malignant leiomyomata have a characteristic appearance on MR. Malignant leiomyomata appear as ill-defined lesions which can be hyperintense on T1 images secondary to hemorrhagic changes. Well-defined lesions are nearly always benign.

MRI can also appreciate different signal intensities to differentiate adenomyosis from an enlarged, myomatous uterus. Different signal intensities within myomas can also be helpful in predicting response to treatment in patients considering treatment with GnRH or uterine artery embolization. Embolization occludes vessels leading to myomas and causes them to regress by cutting off their blood supply. Although submucosal fibroids respond well to embolization, this procedure is still not recommended in women desiring future fertility as there are only limited data available on pregnancy outcomes.

Key Points

After reading this chapter you should be able to understand and identify:

1. The strengths and weaknesses of different diagnostic imaging modalities used in reproductive endocrinology

2. Appropriate clinical situations to utilize different imaging modalities

3. How to perform and interpret hysterosalpingograms and sonohysterograms

4. The appropriate imaging modalities used in the clinical workup for infertility

Uterus Axial Mri
Figure 12.23. Axial MRI of myomatous uterus.

Special thanks to Dr. Fergus Coakley and the Department of Radiology at UCSF for assistance in providing radiographic images.

Suggested Reading

1. Practice committee of the American Society of Reproductive Medicine. Current Evaluation of amenorrhea. Fertil Steril 2004; 82:S33-39, [Excellent text: Highly recommended for a basic understanding of the causes and appropriate diagnostic evaluation of patients with amenorrhea and primary or secondary infertility].

2. Hricak H, Chang YC, Thurnher S. Vagina: Evaluation with MR imaging. Part I. Normal anatomy and congenital anomalies. Radiology 1988; l69(1):l69-74.

3. Lindheim SR, Adsuar N, Kushner DM et al. Sonohysterography: A valuable tool in evaluating the female pelvis. Obstet Gyn Survey 2003; 58:770-84.

4. Ayida G, Chamberlain P, Barlow D et al. Uterine Cavity assessment prior to in vitro fertilization: Comparison of transvaginal scanning, saline contrast hysterosonography and hysteroscopy. Ultrasound Obstet Gynecol 1997; 10:59-62, [Excellent article highlighting the importance of using sonohysterograms to evaluate the endometrial cavity].

5. Soares SR, Barbosa dos Reis MM, Carnagos AF. Diagnostic accuracy of sonohysterography, transvaginal sonography, and hysterosalpingography in patients with uterine cavity diseases. Fertil Steril 2000; 73:406-11.

6. The Rotterdam ESHRE/ASRM-sponsored PCOS consensus workshop group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 81(1):19-25, [Important paper addressing a major condition in reproductive endocrinology].

7. Bansci LF, Broekmans FJ, Eijkemans MJ et al. Predictors of poor ovarian response in in vitro fertilization: A prospective study comparing basal markers of ovarian reserve. Fertil Steril 2002; 77(2):328-36.

8. Bis KG, Vrachliotis TG, Agrawal R et al. Pelvic endometriosis: MR imaging spectrum with laparoscopic correlation and diagnostic pitfalls. Radiographics 1997; 17(3):639-55.

51 Tips for Dealing with Endometriosis

51 Tips for Dealing with Endometriosis

Do you have Endometriosis? Do you think you do, but aren’t sure? Are you having a hard time learning to cope? 51 Tips for Dealing with Endometriosis can help.

Get My Free Ebook


  • tuomas salama
    What happen if hsg spillage block?
    2 years ago
  • Prima
    Can a thin and empty endometrium be corrcet during HSG Xray?
    2 years ago
  • pearl
    How many days can i drink parlodel to corrcet high prolactin?
    2 months ago

Post a comment