A structured system to evaluate urethral support anatomy in magnetic resonance images

It has long been recognized that magnetic resonance (MR) imaging can provide excellent visualization of pelvic floor structures. Several different types of research have been published. Some studies have attempted to determine the optimal MR parameters and technique. Methods Fifty nulliparous women were recruited for an Institutional Review Board-approved case-control study that concerned vaginal delivery and stress urinary incontinence. After the women gave informed consent, multiplanar 2-dimensional fast spin (echo time, 15 ms; repetition time, 4000 ms) proton density MR images of all pelves were obtained by use of a 1.5-tesla superconduct- OBJECTIVE: To develop a systematic method for analyzing the normal magnetic resonance imaging location and appearance of structural features involved in urethral support. STUDY DESIGN: Multiplanar proton density magnetic resonance images of 50 nulliparous women were made at 0.5-cm intervals. The arcuate pubic ligament was the chosen reference point in all views. Structural features were located by identification of the magnetic resonance images in which they appeared. The presence or absence of urethral support structures in each scan level relative to the arcuate pubic ligament was evaluated and recorded as a graphic display. Support structures examined were the arcus tendineus fasciae pelvis, the perineal membrane, the pubococcygeal levator ani muscle and its vaginal and bony attachments, and the pubovesical muscle. Structural definitions were developed on the basis of established periurethral anatomy. Two examiners independently assessed 10 scans for interobserver concordance. This system was used in nulliparous women to quantify the location of magnetic resonance visible structures. RESULTS: Because all levels were 0.5 cm apart, slice identification represented the distance above the arcuate pubic ligament (eg, 2 images above the arcuate pubic ligament or 1.0 cm). Interobserver concordance was 88% in identification of structure location. In the axial plane, specific structural features and relationships were localized. The frequency at which a specific structure was observed at a specific image level in all women was depicted as a gray density within the graphic display. These frequencies indicated where we would expect each structure to be located in healthy women. Relationships between structures and their attachments became apparent. Consideration of expected structural locations proven by nulliparous scans will enable us to quantitatively define abnormal structures in parous women. CONCLUSIONS: This systematic magnetic resonance evaluation allows, for the first time, quantification of the normal anatomic location of urethral support structures. It can be used to help identify the difference between structural abnormalities and normal variation in parous women. (Am J Obstet Gynecol 2001;185:44-50.