Fig. 9  Contrast enhancement of cervical cancer. (a) T2w TSE image in sagittal orientation. The cervical cancer (asterisk) is seen as a large mass of intermediate to high signal intensity that is delineated against a very thin margin of low-signal-intensity cervical stroma and against the more hypointense myometrium of the uterine corpus.

There is barrel-shaped expansion of the cervix. (b) T1w TSE image with FS in sagittal orientation 1 min after Gd-DTPA administration. Enhancement of the hypointense and hypovascularized cervical cancer is inhomogeneous (asterisk) and tumor delineation is not improved by contrast medium administration

abdomen in transverse orientation from the renal hili to the aortic bifurcation (Fig. 11).

In patients with locally advanced or recurrent cervical cancer, it is often necessary to exclude ureteral obstruction, which can be done on transverse T2-weighted images that also serve to assess the para-aortic lymph nodes. In addition, coronal T2-weighted turbo-spin echo (TSE) sequences enable excellent evaluation for possible urinary retention and require little extra time to acquire. In addition, contrast-enhanced MR urography can be performed to exclude tumorinduced hydronephrosis.

Rectouterine or vesicouterine fistulas may develop as a complication of radiochemotherapy. A clinically suspected fistula is an indication for con- trast-enhanced MRI using an unspecific gadolin- ium-based low-molecular contrast medium. Fat-saturated T1-weighted sequences in transverse and sagittal planes are acquired before and after CM administration. The postcontrast sequence is started 60 s after CM administration to image the venous phase. A fistula is typically a filiform structure of

low-signal-intensity lumen showing contrast enhancement of the wall. In addition, T2-weighted inversion recovery sequences that null the signal of fat have a high accuracy in detecting fistulas.

A technical overview of the recommended MR protocol is given in Table 4.

2.2.2\ Dynamic MRI

Dynamic contrast-enhanced MRI of the uterine cervix is performed to evaluate the course of contrast enhancement in a region of interest (ROI) placed in a suspicious area. First, an unenhanced sequence with an acquisition time of about 23 s is acquired, followed by the first postcontrast acquisition about 15–20 s after CM administration. Postcontrast acquisition for measurement of signal intensity in the ROI is repeated over a period of about 10 min. A dynamic MRI study is part of the routine protocol for differentiation of post-thera- peutic changes from recurrent tumor but is rarely necessary in the pretreatment evaluation of patients with cervical cancer. Although vital tumor tissue typically shows earlier arterial enhancement than

Fig. 10  Contrast enhancement of cervical cancer. (a–c) T2w TSE images in sagittal and transverse orientation. Local recurrence of cervical cancer (arrows) is seen as a tumor of intermediate signal intensity above the vaginal

stump. (d–f) T1w TSE images with fat saturation in sagittal and transverse orientation 1 min after Gd-DTPA administration show local recurrence as an enhancing tumor (arrows). Accessory finding in (a): Bartholin cyst (asterisk)

Fig. 11   Illustration of imaging protocol. (a, b) T2w TSE images in sagittal and transverse orientation. (c, d) T1w TSE images with fat saturation in sagittal and transverse orientation. (e, f) T1w TSE images with FS after contrast medium administration in sagittal and transverse orienta-

tion. (g, h) Axial DWI. (i) PD-TSE image for lymph node assessment in the true pelvis. (j) T2w TSE image with respiratory-gating (PACE) for lymph node assessment in the abdomen

the surrounding cervical stroma (Hricak et al. 1991), no benefit was found for differentiation of the tumor or demonstration of parametrial infiltration in studies that evaluated color-coded dynamic MR images (Postema et al. 1998, 1999). However, the demonstration of necrosis and determination of tumor vascularization might help to estimate the radiosensitivity of a tumor prior to therapy. Moreover, a correlation was found between contrast medium time-intensity curves and angiogenic activity as an indicator of infiltration of the lymphatic system (Hawighorst et al. 1998). No correlation was found for other malignancy criteria such as infiltration depth and metastatic pelvic nodes.

2.2.3\ Coil Technique

Morphologic MRI staging of cervical cancer implies high demands on SNR, signal homogeneity, and spatial resolution. Therefore, body phased-array surface coils are recommended because they increase the SNR in comparison to volume coils and can also increase the spatial

resolution, or, alternatively, shorten the acquisition time. The surface coil is placed on the torso and must cover the entire imaging area. Although local coils such as endorectal or endovaginal coils have proven to be more accurate than external array for detecting small cervical cancers, they are rarely used as they distort anatomy, for instance, by compressing lipid lamellae (deSouza et al. 2006). Moreover, an additional examination with a body phased-array coil is required in most cases to evaluate the entire pelvis including the lymph nodes. Since body phased-array coils offer a high SNR, are easy to handle, and provide better patient comfort, local coils have not yet become widely accepted (Fig. 12).

2.2.4\ Vaginal Opacification

Evaluation of vaginal involvement is not the most important goal of MRI staging, as the vagina can be adequately evaluated by clinical examination and colposcopy in most patients. On the other hand, imaging vaginal infiltration is important for