MRI of the Uterus, Cervix, and Vagina


▪ Introduction

Magnetic resonance imaging (MRI) serves as the most comprehensive and conclusive imaging modality available to image the female pelvis. The inherent zonal anatomy of the uterus is exquisitely depicted as a function of the different water content and histology of each mural layer. The predictable MR appearance of the uterus renders identification of abnormalities straightforward ( Fig. 9.1 ). Improved tissue contrast elevates the sensitivity for neoplastic and malignant features compared with other modalities. The most common indications for MRI of the uterus, cervix, and vagina include identification and characterization of leiomyomata, assessment of leiomyoma treatment response, problem solving inconclusive ultrasound or CT findings, and adjunctive staging of malignancies ( Table 9.1 ).

▪ FIG. 9.1, Normal uterine mural stratification. Fat-suppressed sagittal T2-weighted image through the uterus shows the normal trilaminar appearance with the central hyperintense zone (the endometrium— thin arrows ), the middle hypointense zone (the junctional zone— thick arrows ), and the outer isointense zone (outer myometrium— open arrow ).

TABLE 9.1
Female Pelvis Magnetic Resonance Imaging Indications
Clinical Presentation Imaging Objective Details
Pelvic pain Nonspecific Standard protocol
Dysmenorrhea (painful menstruation)
Menorrhagia (>80 mL/cycle)
Metrorrhagia (light blood, irregular intervals)
Menometrorrhagia (>80 mL, irregular intervals)
Exclude endometrial and/or cervical lesions (polyp, leiomyoma, neoplasm)
Exclude adenomyosis or endometriosis
Standard protocol
Abnormal pelvic examination
Delayed menses or precocious puberty
Postmenopausal bleeding
Evaluation of pelvic pain or mass
Exclude endometrial or cervical cancer or polyp Standard protocol
Pain or fever after pelvic surgery or delivery Exclude endometritis or hematoma Review gradient echo images for susceptibility
Localization of intrauterine device Nonanatomic susceptibility artifact (usually linear or curvilinear) Review gradient echo images for susceptibility
Evaluation of infertility
Congenital anomalies
Müllerian duct anomalies ± Dedicated imaging planes
Uterine leiomyoma evaluation Location (submucosal vs intramural vs subserosal), vascularity, degeneration (cystic vs hemorrhagic) ± Dedicated imaging planes
Assessment for pelvic floor defects Cystocele, enterocele, vaginocele, rectocele, pelvic floor descent ± Dynamic maneuvers
Clarification of indeterminate imaging findings Follow up previously detected abnormality (ie, hemorrhagic cyst)
Further characterization of abnormality detected in another imaging study
Standard protocol
Known or risk of malignancy Screening for malignancy in patients with increased risk
Detection and staging of gynecologic malignancy
Tumor recurrence assessment
Presurgical/laparoscopic evaluation
Consider vaginal gel if cervical or vaginal involvement

Technique

Given the potentially small size of uterine, cervical, and vaginal lesions and the need for high resolution imaging, the examination justifies the use of a high field strength system (≥1.0 T). Nonetheless, diagnostic images can be obtained on systems of lower field strength ( Fig. 9.2 ). A dedicated phased array coil guarantees optimal signal to noise. An antiperistaltic agent (such as glucagon or hyoscyamine) may be used to eliminate motion artifacts from bowel activity. The application of vaginal gel may also be considered to facilitate evaluating vaginal and cervical lesions.

▪ FIG. 9.2, 0.3-Tesla images. (A) Sagittal T2-weighted image through the uterus obtained using a 0.3-Tesla system reveals an isointense endometrial mass with no deep myometrial invasion, corresponding to endometrial carcinoma. (B) The corresponding axial T2-weighted image shows normal endometrium (thin arrow) abutting the distal aspect of the mass and an intramural fibroid (thick arrow) . Sagittal ( C ) and axial ( D ) T2-weighted images through the uterus obtained using a 0.3-Tesla system in a different patient show a tubular mildly hyperintense lesion with a hypointense core within the endometrial cavity found to be an endometrial polyp. Also note the partially subserosal fibroid arising from the posterior uterine body ( arrow in D ).

Start off viewing a localizer sequence with a large field of view to assess coil placement (ensuring maximal signal emanating from the region of interest and not the upper thighs or lower abdomen). Single-shot fast spin echo T2-weighted images or balanced gradient echo sequences yield the most diagnostic information. Configure the localizer sequence to include the kidneys because of the association of renal anomalies with Müllerian duct (fallopian tubes, uterus, and proximal vagina) anomalies. The remainder of the examination demands a focused approach with a higher spatial resolution and a field of view in the range of 24 cm.

A combination of T1-weighted, T2-weighted, and fat-saturated sequences suffices to solve most problems encountered in the pelvis ( Table 9.2 ). T2-weighted images are the mainstay of uterine, cervical, and vaginal imaging. T2-weighted images display the trilaminar anatomy of the uterus—the endometrium, the inner myometrium (junctional zone), and the outer myometrium ( Fig. 9.1 ), which are contiguous with the endocervical glands, the fibrous stroma of the endocervix, and the looser connective tissue of the ectocervix. Oblique coronal and axial images orthogonal to the axis of the uterus supplement the examination to characterize potential Müllerian duct anomalies, if suspected ( Figs. 9.3 and 9.4 ). T1-weighted images depict hemorrhage and lipid and the addition of fat-saturated sequences allows for differentiating between blood and fat (which both appear hyperintense in T1-weighted images without fat suppression). In- and out-of-phase images serve as a time-saving alternative to conventional spin-echo T1-weighted images (approximately 20 seconds compared with 3–5 minutes) with sensitivity to both intracellular fat and susceptibility artifact, although potentially degraded by low signal-to-noise on low field strength systems.

TABLE 9.2
Female Pelvis Magnetic Resonance Imaging Protocol
Pulse Sequence Details (TR/TE) Field of View (cm) Slice Thickness (mm)
Coronal localizing sequence SSFSE or HASTE (5000/180)
Balanced gradient echo (min/min)
32 5
Axial T2 ± Fat suppression (4000/100) 24 5
Axial T1 In-/out-of-phase (120/2.2, 4.4)
FSE (600/10)
24 5
Sagittal T2 ± Fat suppression (4000/100) 24 5
Dynamic contrast Axial or sagittal 3D GRE with fat suppression (min/min) 24 4–5 with 50% overlap
Delayed postcontrast Axial, sagittal, and coronal 3D GRE with fat-suppression (min/min) 24 4–5 with 50% overlap
Diffusion EPI (3000/60) 24 5–8 skip 0–1

▪ FIG. 9.3, Oblique coronal imaging of the uterus. The full extent of a fibrous septum (arrow) of a septate uterus (and other congenital anomalies) is well delineated by orienting the imaging plane along the long axis of the uterus to obtain an oblique coronal image.

▪ FIG. 9.4, T-shaped uterus. (A) Obliquely coronally reformatted T2-weighted image in a 47-year-old woman with a history of fetal diethylstilbestrol exposure elegantly portrays the aberrant anatomy that is less clearly rendered with coronal (B) , sagittal (C) , and axial (D) planes prescribed orthogonally to the axes of the body.

Gadolinium-enhanced images provide additional information regarding the complexity and/or blood supply of a lesion and its tissue content and often increase lesion conspicuity. Dynamic imaging provides a reliable time frame to assess enhancement patterns, whereas static pre and postgadolinium images yield only binary information (enhancement versus no enhancement). Quantitative evaluation of leiomyoma vascularity demands dynamic images.

Three-dimensional fat-saturated gradient echo images offer the best spatial resolution and tissue contrast. 0.1 mmol/kg is administered intravenously at approximately 1 to 2 mL/second. A timing bolus or timing sequence triggers the arterial phase of the acquisition and one or two additional phases obtained in succession suffice. A delayed T1-weighted (preferably fat-saturated) sequence detects delayed enhancement, if present.

Interpretation

If pictures are worth a thousand words, imagine how much information the hundreds of images in the multiimage sets of a female pelvis MR study are worth. This volume of information compels the use of a directed search pattern ( Box 9.1 ). First of all, assess the technical adequacy of the examination. Review the localizer sequence, which usually includes large field-of-view coronal images, and ensure that coil placement is adequately reflected by the highest signal emanating from the region of interest (and not the abdomen or subpelvic region) ( Fig. 9.5 ). Note whether gadolinium was administered and whether enhancement is perceptible. Assess the degree of motion artifact and any other artifact that degrades image quality ( Fig. 9.6 ).

BOX 9.1
Female Pelvis Checklist

Technical

  • Magnetic field strength

  • Coil position

    • Signal–optimal signal corresponding to pelvis

    • Kidneys–at least one large FOV coronal to include kidneys

  • Enhancement–note amount and type of contrast agent

    • Inspect vessels for adequate enhancement

  • Artifacts

    • Bowel peristalsis

    • Susceptibility–surgical hardware, gas (ie, bowel)

    • Motion–bulk motion, vascular flow artifacts

    • Conductivity/dielectric effects – focal signal loss

Uterus

  • Size

  • Position–version and flexion

  • Endometrium

    • Thickness

    • Homogeneity

    • Focal lesions

    • Intraluminal fluid or susceptibility (eg, gas or IUD)

  • Inner myometrium (junctional zone)

    • Thickness

    • Sharpness of border

    • Intramyometrial hyperintensities

  • Cesarean section defect

  • Leiomyomata–size, location, degeneration, and vascularity

Cervix

  • Mucosa

    • Nabothian cysts

    • Thickness

    • Luminal fluid

    • Stroma

    • Integrity

    • Parametrial infiltration

Vagina

  • Cystic lesions (upper vs lower)

  • Vaginal wall

    • Focal vs diffuse thickening

    • Invasion from adjacent structures

  • Tampon

Other Anatomy

  • Bladder

  • Bowel

  • Musculoskeletal structures

    • Bony pelvis

    • Lower lumbar spine

    • Muscles–gluteal, adductors, hip flexors, piriformis

    • Tendons–iliopsoas, rectus femoris, hamstring

▪ FIG. 9.5, Poor coil placement. (A) and ( B) Coronal T2-weighted images of the pelvis show maximal signal centered at the level of the pubic symphysis rather than the region of interest—the uterus and adnexa—and remember that the kidneys should be visualized in all cases. Note the homogeneously hyperintense mass in the right hemipelvis in this patient with pelvic lymphoma (arrow) . (C) Coronal T2-weighted image in a different patient demonstrates maximal signal arising from the region of the uterus with visualization of the kidneys (thin arrows) in a patient with cervical carcinoma (thick arrow) . (D) The kidneys (thin arrows) are conspicuously well visualized in this coronal T2-weighted image with maximal signal emanating from the abdominopelvic junction in this patient with a right-sided cystadenofibroma (thick arrow) .

▪ FIG. 9.6, Image degradation as a result of artifacts. Coronal T2-weighted ( A ) and axial T1-weighted fat-saturated gradient echo (B) images depict wraparound artifact along the phase encoding axis ( arrows in A and B ), which do not obscure underlying anatomy and reflect selection of a small field of view, resulting in the aliasing of subcutaneous fat. (C) Sagittal T2-weighted image mildly degraded by breathing motion artifact along the phase encoding axis (arrow) . (D) A combination of susceptibility artifact (thin arrows) and failure of fat suppression (thick arrow) degrades this sagittal T2-weighted fat saturated image in a patient with a focal fundal adenomyoma (open arrow) . (E–G) Artifact arising from bowel peristalsis blurs and obscures normal pelvic anatomy with no evidence of gross patient motion.

Look at the uterus keeping in mind the age and menstrual status of the patient and any relevant history, such as endometrial or cervical carcinoma, cesarean section, or leiomyomata. Measuring the uterus in three orthogonal planes is standard and helps you objectively assess overall uterine size. Observe uterine zonal anatomy (central endometrium, middle junctional zone, and outer myometrium), and measure the thickness of each. Identify any leiomyomata or other uterine lesions and record sizes. Comment on whether the uterus is ante- or retroverted or ante- or retroflexed. In the setting of hysterectomy, record the presence and status of the vaginal cuff and residual cervical or uterine tissue. Look for susceptibility artifact on gradient echo images corresponding to surgical clips, if present.

Confirm the integrity of the fibrous stroma of the endocervix, and exclude the presence of any cystic or solid cervical lesions. Inspect the vagina and vulva, and keep in mind the prevalence of benign developmental and acquired cystic lesions. Exclude focal or diffuse vaginal wall thickening, and note intraluminal fluid if present.

Assess the quantity of free fluid in the pelvis, and remember that a small quantity is physiologic in reproductive-age females. Look for pelvic lymph nodes, and record any enlarged nodes.

Although the bladder is often not optimally evaluated because of incomplete distention (usually patients are instructed to void in order to promote comfort and obviate motion for the duration of examination), do not ignore it. Observe any focal lesions, filling defects/stones, wall thickening, or diverticula. Check the urethra for diverticula.

Trace the bowel from the anus proximally as far as possible. Look at the coronal images, which are often performed with the largest field of view to visualize as much of the bowel and peritoneal cavity as possible. View sagittal and coronal images to assess for pelvic floor laxity.

Finally, use T1-weighted and fluid sensitive sequences to exclude osseous lesions. Sagittal images are useful to detect disc pathology in the lower lumbar spine. Evaluate muscles and tendons (such as the gluteal, adductor, and hip flexor muscles, and iliopsoas, rectus femoris, and hamstring tendons on T2-weighted axial and coronal sequences.

▪ Uterus

Normal Features

In colloquial medical parlance, “uterus” signifies the uterine body specifically or the corpus and “cervix” refers to the uterine cervix. Before you can intelligently comment on the status of the (body of the) uterus you need to know the age and menstrual status of the patient and any relevant surgical history. The size of the uterus is a function of age and reproductive/menstrual history ( Box 9.2 ). Without the stimulating effects of female hormones before puberty, cervical stature exceeds uterine size. Measure the uterus in three orthogonal dimensions along its axis, and comment on uterine positioning (version or flexion) ( Fig. 9.7 ).

BOX 9.2
Normal Dimensions of the Uterus

  • Uterus reproductive age, 8 × 5 cm

  • Uterus perimenarchal or postmenopausal, 5 × 2 cm

  • Endometrium

    • Proliferative 3–8 mm

    • Secretory 5–16 mm

    • Postmenopausal (no bleeding) ≤8 mm

    • Postmenopausal (bleeding) ≤5 mm

▪ FIG. 9.7, Uterine measurement technique and assessment of anteversion/retroversion/anteflexure/retroflexure. Sagittal (A) and axial (B) T2-weighted images of a retroverted uterus illustrate the measurement technique for obtaining the longitudinal (solid line) , height (dashed line) , and width (dotted line) measurements. Note the focal adenomyoma (arrow) .

Try to appreciate the zonal anatomy or mural stratification of the uterus, which is most developed in reproductive-age females ( Fig. 9.8 ; see also Fig. 9.1 ) and fades during menopause. Endometrial cyclical changes only occur in menstruating females; in pre- and postmenopausal females, the endometrial changes are only incurred by pathologic or iatrogenic phenomena and normally measures up to 4 mm in maximal thickness. In menstruating females, endometrial thickness is variable. From the proliferative phase, the endometrium thickens from a minimum of 3 to 8 mm to 5 to 16 mm during the secretory phase. Notwithstanding age and menstrual status, normal endometrium demonstrates uniformly homogeneously near fluid hyperintensity on T2-weighted images.

▪ FIG. 9.8, Zonal anatomy/mural stratification of the uterus.

If the appearance and/or thickness fall outside of the normal range, try to characterize the abnormality as intracavitary fluid or gas, versus a focal or diffuse process. Fluid is a frequent and generally nonpathologic finding in menstruating females. Prepubertal endometrial fluid generally indicates an obstructing lesion (hydrometrocolpos or hydrocolpos). Postmenopausal patients with endometrial fluid often harbor an underlying endometrial or cervical lesion (such as endometrial atrophy, hyperplasia, polyp or carcinoma, and cervical stenosis), but fluid has not been conclusively proven to be pathologic. Gas may be present postprocedurally, postpartum, or in the context of infection (endometritis). Gas induces susceptibility artifact and is most conspicuous on gradient-echo sequences ( Fig. 9.9 ).

▪ FIG. 9.9, Intrauterine gas. Sagittal T2-weighted (A) , axial out-of-phase (B), and in-phase (C) images showing gas within a dehiscent cesarean section defect. Note the blooming in the in-phase gradient recalled echo image (arrow) .

Endometrial Pathology

Diffuse Abnormalities

Diffuse endometrial abnormalities include predominantly endocrinologic/proliferative, infectious, iatrogenic, and neoplastic etiologies ( Box 9.3 ). Global alteration in thickness or homogeneity of the T2 hyperintense endometrial layer is the common denominator. Endometritis falls into two main categories—postpartum and nonpostpartum. Endometritis most commonly follows vaginal delivery, especially with prolonged rupture of membranes, chorioamnionitis, prolonged labor, and retained products of conception. Risk factors for nonpostpartum endometritis include uterine artery embolization, venereal disease, and presence of an intrauterine device. Although endometritis usually relies on clinical findings for diagnosis, imaging studies exclude additional abnormalities in patients with refractory symptoms. Typical findings include diffuse uterine enlargement, intracavitary gas and (often complex) fluid, and a thickened, heterogeneous endometrium. Look for edematous, relatively hypovascular foci subjacent to the abnormal endometrium.

BOX 9.3
Diffuse Endometrial Abnormalities

Endocrinologic/proliferative

  • Polycystic ovarian syndrome

  • Pregnancy

  • Obesity

Infectious

  • Postpartum endometritis

  • Nonpostpartum endometritis

Iatrogenic

  • Exogenous estrogen

  • Tamoxifen

Neoplastic

  • Endometrial carcinoma

  • Cervical carcinoma

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here