In order to optimize pelvic imaging, several common approaches are used. Patients are asked to fast for approximately 4 hours beforehand, with the use of antispasmodic agents recommended in order to further reduce peristalsis-related artifact. Patients are additionally asked to empty their bladders immediately before the study and scanning is performed with patients in a supine position. Endovaginal contrast can be used to assess for vaginal
invasion from cervical or endometrial tumors; however, because this can be adequately assessed clinically, its use is not widespread. Endorectal or endovaginal coils are generally not recommended; although they offer an increased signal-to-noise ratio, they have the disadvantage of distorting the local anatomy and are invasive. Moreover, due to their high sensitivity, any motion of the coil will lead to considerable ghosting artifacts. Use of multichannel phased array coil (pelvic or cardiac) is standard to decrease scan time and improve image resolution of the anatomical sequences.⁶ These coils also allow for large field-of-view sequences to evaluate pelvic lymph nodes.

Three-dimensional spoiled steady state free precession gradient-echo T₁-weighted images are acquired at different time points following injection of gadolinium (0.1 mmol/kg body weight) via a power injector immediately followed by saline flush. For multiphase MRI, three or four time points are selected, with the imaging plane determined by the tumor to be imaged (i.e., sagittal planes for endometrial and cervical tumors and axial planes for ovarian lesions), with a further delayed phase in a second plane (Table 61.1 and 61.2). For DCE-MRI, series are acquired at multiple consecutive time points. Individually selected AIFs are typically derived from the larger pelvic vessels: the external iliac or femoral arteries. Nevertheless, these vessels are still of relatively small calibre and prone to partial volume effects from pulsation, which may be confounded by inflow effects.^2,7 It is possible to use model AIFs from population-averaged data in order to overcome these issues,^1,8,9,10 however, this is somewhat counterproductive because these AIFs do not account for individual differences in cardiac output, bolus arrival, total blood volume, and so forth. Although its use is not universally recommended for DCE-MRI, parallel imaging is a potential means of achieving improved spatial coverage and temporal resolution to allow a larger field of view for DCE-MRI to assess for treatment response in more remote deposits.²

Multiphase MRI for pelvic imaging has generally been adopted as part of standard practice. The advantages over DCE MRI are the reduced time taken to acquire, the lack of postprocessing, and the relative ease of qualitative assessment. DCE MRI is generally only used in larger institutions and in certain problem-solving situations, vide infra.

Ultrasound is the first-line imaging test for assessment of adnexal lesions, with MRI playing a secondary role in problem solving and further characterizing complex adnexal masses as being benign, borderline, or malignant and helping to provide prognostic information and inform treatment decisions.¹¹ Standard anatomical MR sequences can confidently characterize benign lesions such as simple or hemorrhagic cysts, endometriomas, and mature cystic teratomas. Fat-saturated T₁WI helps to distinguish between hemorrhage and fat within a lesion. Typically, endometriomas are hyperintense on T₁WI due to the presence of methemoglobin, retaining this high signal on fat-saturated T₁WI, and are of low to intermediate signal intensity on T₂WI due to the presence of iron from blood degradation products. Mature cystic teratomas can be confidently diagnosed when fat is demonstrated within a cystic lesion,³ which appears as high signal intensity on T₁WI and T₂WI, with a drop in signal intensity following fat suppression (Fig. 61.1).

Morphologic features indicative of ovarian malignancy include size greater than 4 cm, presence of necrosis within a solid lesion, the presence of solid components or vegetations or papillary projections within a cystic lesion, and wall or septal thickening greater than 3 mm. Presence of at least one of these primary features and one ancillary finding of either involvement of pelvic organs,


ascites, peritoneal disease, or lymphadenopathy has been shown to correctly characterize malignancy in 95% of cases (Table 61.3).¹² Transvaginal ultrasound and MR have similar sensitivity for the detection of solid components within cystic masses, but MR has improved accuracy due to its increased specificity.¹³

Accurately characterizing complex cystic adnexal masses as benign, borderline, or malignant can prove challenging and postcontrast imaging often adds value.¹³ Multiphase contrast-enhanced MRI can add additional information such as helping to differentiate solid components or papillary projections from clots and debris.¹⁴ Contrast-enhanced MRI also improves confidence for the detection of solid components in a cystic lesion (Fig. 61.2) and necrosis in a solid lesion by increasing conspicuity.¹³ In addition, early enhancement at multiphase MR has been shown to be a feature indicative of borderline malignant tumors and can help to differentiate them from benign ovarian tumors.^12,15,16,17 Contrast-enhanced images improve detection of peritoneal implants, which appear as nodular or plaque-like soft tissue masses on the delayed (5-minute postgadolinium) series.¹⁸ Thus, contrast-enhanced MR can help improve planning of cytoreductive chemotherapy in advanced primary or relapsed ovarian carcinoma.

Ovarian cancer carries a generally poor prognosis, mainly due to the typically advanced stage at presentation, but also in part due to high rate of residual or recurrent disease.¹⁹ A mixed response frequently occurs at different abdominal sites of ovarian metastases. This can be explained, in part, by the variations in blood supply and consequent hypoxia, which limit delivery and efficacy of chemotherapy, or by a more desmoplastic response in high-grade serous tumors. DCE MRI can therefore be useful in the assessment of such vascular variations and in predicting treatment response. Indeed, DCE MRI has been shown to be predictive of future progression in asymptomatic patients with advanced ovarian cancer as early as 4 weeks following completion of chemotherapy (Fig. 61.3).²⁰