Share this article on:

Magnetic Resonance Imaging of Normal Levator Ani Anatomy and Function

Singh, Kavita MRCOG; Reid, W. M. N. MRCOG; Berger, L. A. FRCR

Original Research

OBJECTIVE: To evaluate the anatomy and function of the levator ani in normal women by dynamic magnetic resonance imaging.

METHODS: Twelve asymptomatic, nulliparous, premenopausal women with no previous pelvic surgery underwent a dynamic magnetic resonance imaging scan of their pelvis. The origin, orientation, thickness, and function of the two components of the levator ani were studied.

RESULTS: The ileococcygeus is a thin muscle with an upward convexity. It slopes forward and medially. It is of variable thickness (mean thickness 2.9 mm, standard deviation 0.8 mm). There are apparent gaps in the muscle diaphragm and at its site of origin from the obturator fascia. The puborectalis is a thicker muscle. It is shaped like a belt encasing the pelvic organs. It is taller posteriorly than anteriorly. It is not attached to the bladder neck, but the midurethra and lower urethra lie in close proximity to it. The puborectalis moves dorsoventrally, whereas the ileococcygeus moves craniocaudally.

CONCLUSION: The levator ani is not a single muscle but has two functional components that vary in thickness, origin, and function. The ileococcygeus has a mainly supportive function, whereas the puborectalis has a sphincteric function. Gaps in the diaphragmatic portion of the ileococcygeus are a normal finding. Individual components of the levator ani may be prone to different types of childbirth trauma and should therefore be assessed separately when planning rehabilitation.

Anatomy and function of levator ani is studied in normal nulliparous women using dynamic magnetic resonance imaging.

Department of Obstetrics and Gynaecology and Department of Radiology, Royal Free and University College Medical School, London, United Kingdom.

Address reprint requests to: Kavita Singh, MRCOG, Royal Free and University College Medical School, Department of Obstetrics and Gynaecology, Rowland Hill, London NW3 2PF, United Kingdom; E‐mail:

Received January 30, 2001. Received in revised form April 26, 2001. Accepted June 15, 2001.

The bipedal gait in Homo sapiens causes problems of support of the internal organs, which may be exacerbated by childbirth. The levator ani muscles play an important role in supporting the pelvic organs and maintaining their continence.1–3 To better understand the changes these muscles undergo in pelvic floor prolapse, it is essential to study their anatomy in normal women. The pelvic floor muscles lie hidden in the pelvis and are difficult to assess clinically.4 Multiplanar imaging with magnetic resonance imaging (MRI) permits this detailed study. Static images show their morphology. Dynamic images show the functional changes that occur on straining and contraction of the levator ani. We conducted this study to clarify muscle morphology and function and its role in supporting the pelvic floor.

Back to Top | Article Outline


Verbal consent was obtained from 12 female volunteers from the staff of our hospital; participants also signed a consent form before having their MRI scans. The Royal Free Hospital ethical committee approved the study. All participants were nulliparous, premenopausal women (age range 23–42 years) with no previous pelvic surgery. They had no history of urinary or bowel incontinence nor symptoms of prolapse. No other factors such as constipation, previous hernia repair, varicose veins, or chronic cough were present. A vaginal examination was not performed, but they were taught the Valsalva maneuver.

All images were performed with the women lying supine. Static T2 TSE images were obtained using a 1.5 Tesla Philips Gyroscan at 6 mm/6 mm on coronal section and 5 mm/5 mm on axial and sagittal sections, with a field of view of 280, matrix 230 × 512, TR 6086/TE 150, 4 NEX, and 4.5 minutes acquisition time per plane. Dynamic MRI was performed using a fast spin echo sequence. The slices were obtained at 5 mm/5 mm, with a field of view of 350, matrix 138 × 256, TR 8633/TE 80, 2NEX, and 17 seconds acquisition time per plane. Coronal, axial, and sagittal images at rest and on maximal straining were analyzed on a workstation (Easyvision) with zoom facility and electronic calipers. The origin, orientation, thickness, and changes of the levator ani on performing the Valsalva maneuver were studied.

The ileococcygeus was demonstrated best on the coronal and sagittal images. Its origin, thickness, and orientation were measured on the coronal sections (Figure 1A, 1B, 1C). Its orientation was analyzed by assessing its slope. We measured the angle between the ileococcygeus and the transverse plane of the pelvis on serial coronal sections (Figure 1B). The ileococcygeal angle was measured at 1‐cm intervals starting at the level of vagina and proceeding posteriorly for 5 cm. The transverse plane of the pelvis was obtained by joining corresponding bony landmarks on the pelvic sidewall. This method corrected for any pelvic rotation. We joined the upper edge of the femoral head on the anterior slices and the upper end of the ischium on posterior slices. Changes in the slope of the ileococcygeus were studied both at rest and on straining to assess movement. The puborectalis was studied on axial and sagittal sections. We measured its height on the sagittal sections (Figure 2). The levator hiatus was also measured on sagittal films as the distance between the pubis and the outer border of the puborectalis. Any changes in the size of the levator hiatus were noted on straining (Figure 2B). The thickness of the puborectalis was measured on axial images in its anterior third (Figure 3) at the level of the inferior border of the pubic symphysis.

The Minitab statistical package was used to analyze the data. The results are given as mean values with standard deviation (SD). The Wilcoxon signed rank test was used to measure the differences in all the data at rest and during maximum straining. Statistical significance was assigned to a P value of <.05.

Back to Top | Article Outline


The levator ani consists of two parts: 1) the inferomedial part or puborectalis, and 2) the superolateral part or ileococcygeus (Figure 1A). The origin of the ileococcygeus was clearly visible from the fascia overlying the obturator internus on coronal and sagittal images. Fibro‐fatty tissue separates the muscle bundles at their origin, which can appear as gaps on coronal sections (Figure 1A). These gaps were also noted in the muscle diaphragm on sagittal sections in all 12 women (Figure 2A). The mean thickness of the ileococcygeus muscle measured at rest on coronal sections at the level of the ischial tuberosity was 2.8 mm (SD 0.80 mm) (Table 1). On coronal sections, the muscle has a medial slant and a cranial convexity. The slope of the ileococcygeus, as measured by the ileococcygeal angle, decreased progressively as one moved from anterior to posterior (Figure 4). On straining, there was no uniform pattern of change in the ileococcygeus. In seven women, there was elevation of the ileococcygeal diaphragm on straining, and in five, there was descent of the ileococcygeal diaphragm. The posterior part descended more than the anterior part. On straining, there was a significant increase in muscle thickness (P = .003) (Table 1, Figure 5).

The puborectalis was thicker than the ileococcygeus and formed a band around the urethra, vagina, and rectum. It was seen to arise from the lower lateral border of the pubic symphysis. No attachments to the bladder neck were seen, but the mid‐ and lower urethra lay in close proximity to its anterior portion. The puborectalis acts like a belt encasing the pelvic organs. It is taller posteriorly than anteriorly. The muscle moves dorsoventrally and narrows the levator hiatus on straining, although this narrowing is not statistically significant (P = 287). The right side was often thinner than the left. The mean thickness of the right puborectalis was 4.9 mm (SD 2.3), and the left side was 6.5 mm (SD 2.04) (P value for the difference between right and left side thickness was .003).

Back to Top | Article Outline


Strohbehn et al5 have described the components of the levator ani in detail by scanning two female cadavers with MRI. Our study of 12 healthy nulliparous live women supports their observations. Traditionally, the levator ani has been divided into three parts—the pubococcygeus, the ileococcygeus, and the puborectalis, based on their site of origin and insertion.6 Our observations using dynamic MRI show that the levator ani consists of two components—an inferomedial part, the puborectalis, and a superolateral part, the ileococcygeus. They lie in different planes in the pelvis and have different functions. The inferomedial part encircles the lower rectum and the anal canal and functions as a “puborectoanalis.” It is responsible for maintaining the angulation of the anorectal junction and contributes to anal continence. The superolateral part of the levator ani has a supportive role. It is a single sheet of muscle that arises from the pubis and the ileum and inserts onto the coccyx. It functions as a “puboileococcygeus.”

The levator ani has been described as having a forward and medial slope. This is said to be responsible for rotating the fetal head internally in normal labor. Hugosson et al7 and Hjartardottir et al8 highlighted the dome shape of the levator diaphragm but disregarded its downward forward slant. Their observations, however, were made on a coronal scan only. Our observations of the change in the slope of the ileococcygeus at different levels show that it is a dynamic muscle and that it is convex cranially and has a forward and medial slope. This results naturally from its higher level of origin and lower level of insertion (decussation) onto the levator plate. As it is a thin muscle, it is difficult to explain how this cranially convex shape is maintained. In slightly over half of our volunteers, the ileococcygeus elevates on straining in accordance with Zacharin's9 observations. Gaps containing connective tissue and fat were noted in the diaphragmatic portion of the ileococcygeus and at its site of origin. They reflect the thinness of the muscle bundles.

It is important to examine all three planes on MRI as different information is derived from each. We found the axial plane to be the most useful for studying the puborectalis and the coronal plane for studying the ileococcygeus. The midsagittal plane showed the ileococcygeal raphe (levator plate) and the puborectalis best, and the parasagittal sections the origin of the ileococcygeus.

A number of studies have looked at the role of MRI in the evaluation of the pelvic floor muscles.10–12 Ozasa et al11 studied the integrity of the levator plate as a predictor of uterovaginal support. They found that when a line extrapolated from the levator plate on sagittal section crossed the pubis, it excluded prolapse. As the levator plate is formed by the ileoccygeus, we hypothesize that women with a weak ileococcygeus develop vaginal prolapse, whereas women with a weak puborectalis develop problems with incontinence. Bo et al12 studied the movements of the coccyx and the bladder neck on midsagittal MRI sections and implied that they reflected the movements of the levator ani. They believed there was a concentric movement of the levator ani lifting the coccyx upwards and ventrally. Kegel13 also described the way that the pelvic floor muscles contract as being a combination of a squeeze and an inward lift. This reflects the multicomponent action of the levator ani where the puborectalis provides the inward squeeze and the ileococcygeus the upward lift.

Asymmetry of the puborectalis was observed by Fielding et al14 and Tunn et al.15 We also have observed a statistically significant thinning of the right puborectalis (P = .003) (Figure 6). Tunn et al15 have attributed this difference to the chemical shift artefact. Reverse phase encoding in our group did not significantly change this difference, which we believe cannot be explained by chemical shift alone, although the number of participants in our study was small.

Goh et al16 used dynamic MRI to assess the descent of pelvic organs in asymptomatic subjects and found unsuspected low‐grade prolapse in 14% (seven of 50) of these women. Our results suggest a possible explanation. The puborectalis lies in a lower plane and is more likely to undergo damage during the second stage of labor, whereas the ileococcygeus lies on a higher plane. It is more prone to damage during the first stage of labor or from any condition associated with increased intra‐abdominal pressure such as obesity and constipation. Further detailed study of the levator ani in affected women with asymptomatic prolapse would be interesting as the degree of damage may initially be asymptomatic.

Dynamic MRI makes it possible to study the anatomic variations and functional changes in the levator ani muscle. Assessing the variations of this muscle in normal women would improve the understanding regarding the morphologic changes in the levator ani of women with pelvic floor disorders. MRI is useful for defining the anatomic configuration of the levator ani and can assess its function on dynamic imaging.

Back to Top | Article Outline


1. Kirschner-Hermanns R, Wein B, Niehaus S, Schaefer W, Jakse G. The contribution of magnetic resonance imaging of the pelvic floor to the understanding of urinary incontinence. Br J Urol 1993;72:715–8.
2. Delancey JOL, Starr RA. Histology of the connection between the vagina and levator ani muscles: Implications for the urinary function. J Reprod Med 1990;35:765–71.
3. Klutke C, Golomb J, Barbaric Z, Raz S. The anatomy of stress incontinence: Magnetic resonance imaging of the female bladder neck and urethra. J Urol 1990;143:563–6.
4. Wall LL. The muscles of the pelvic floor. Clin Obstet Gynecol 1993;36:910–23.
5. Strohbehn K, Ellis JH, Strohbehn JA, DeLancey JO. Magnetic resonance imaging of the levator ani with anatomic correlation. Obstet Gynecol 1996;87:277–85.
6. Bustami FMF. A reappraisal of the anatomy of the levator ani muscle in man. Acta Morphol Neerl Scand 1988;26:255–68.
7. Hugosson C, Jorulf H, Lingman G, Jacobsson B. Morphology of the pelvic floor. Lancet 1991;337:367.
8. Hjartardottir S, Nilsson J, Peterson C, Lingman G. The female pelvic floor: A dome — not a basin. Acta Obstet Gynecol Scand 1997;76:567–71.
9. Zacharin RF. Pulsion enterocele: Review of functional anatomy of the pelvic floor. Obstet Gynecol 1980;55:135–40.
10. Plattner V, Leborgne J, Heloury Y, Cohen JY, Rogez JM, Lehur PA, et al. MRI evaluation of the levator ani muscle: Anatomic correlation and practical applications. Surg Radiol Anat 1991;13:129–31.
11. Ozasa H, Mori T, Togashi K. Study of uterine prolapse by magnetic resonance imaging: Topographical changes involving the levator ani muscle and the vagina. Gynecol Obstet Invest 1992;34:43–8.
12. Bo K, Lilleas F, Talseth T, Hedland H. Dynamic MRI of the pelvic floor muscles in the upright sitting position. Neurourol Urodyn 2000;20:167–74.
13. Kegel AH. Stress incontinence and genital relaxation. Ciba Clin Symp 1952;2:35–51.
14. Fielding JR, Dumanli H, Schreyer AG, Okuda S, Gering DT, Zou KH, et al. MR-based three-dimensional modelling of the normal pelvic floor in women: Quantification of muscle mass. Am J Roentgenol 2000;174:657–60.
15. Tunn R, Paris S, Fischer W, Hamm B, Kuchinke J. Static magnetic resonance imaging of the pelvic floor muscle morphology in women with stress urinary incontinence and pelvic prolapse. Neurourol Urodyn 1998;17:579–89.
16. Goh V, Halligan S, Kaplan G, Healy JC, Bartram CI. Dynamic MRI imaging of the pelvic floor in asymptomatic subjects. Am J Roentgenol 2000;174:661–6.
© 2002 The American College of Obstetricians and Gynecologists