Research www. AJOG.

org

BASIC SCIENCE: GYNECOLOGY

Perineal body anatomy in living women: 3-dimensional
analysis using thin-slice magnetic resonance imaging
Kindra A. Larson, MD; Aisha Yousuf, MD; Christina Lewicky-Gaupp, MD; Dee E. Fenner, MD; John O. L. DeLancey, MD

OBJECTIVE: The objective of the study was to describe a framework for
visualizing the perineal body’s complex anatomy using thin-slice mag-
netic resonance (MR) imaging.
STUDY DESIGN: Two millimeter thick MR images were acquired in 11
women with normal pelvic support and no incontinence/prolapse symp-
toms. Anatomic structures were analyzed in axial, sagittal, and coronal
slices. Three-dimensional (3-D) models were generated from these
images.
RESULTS: Three distinct perineal body regions are visible on MR imag-
ing: (1) a superficial region at the level of the vestibular bulb, (2) a midre-
gion at the proximal end of the superficial transverse perineal muscle,
and (3) a deep region at the level of the midurethra and puborectalis
muscle. Structures are best visualized on axial scans, whereas cranio-
caudal relationships are appreciated on sagittal scans. The 3-D model
further clarifies interrelationships.
CONCLUSION: Advances in MR technology allow visualization of peri-
neal body anatomy in living women and development of 3-D models
that enhance our understanding of its 3 different regions: superficial,
mid, and deep.
Key words: magnetic resonance imaging, pelvic floor anatomy,
perineal body, posterior wall

Cite this article as: Larson KA, Yousuf A, Lewicky-Gaupp C, et al. Perineal body anatomy in living women: 3-dimensional analysis using thin-slice magnetic
resonance imaging. Am J Obstet Gynecol 2010;203:494.e15-21.

T he perineal body is an anatomical

structure encountered daily by ob-
stetricians and gynecologists. Known to
Understanding the relationships of its
components is crucial for successful ob-
stetrical laceration and fistula repairs in
University of Michigan Institutional
Review Board–approved (#1999-0395)
case-control study of pelvic organ pro-
some as the anchor of the pelvis, its com- addition to other procedures such as lapse. Women in the control group were
ponents have long been debated in the perineocele repair. In this study, we asymptomatic based on Pelvic Floor Dis-
literature in cadaver dissections and his- sought to utilize advanced thin-slice tress Inventory (PFDI) and Pelvic Floor
tological studies.1-7 Like the word shoul- magnetic resonance imaging (MRI) of Impact Questionnaires (PFIQ), had neg-
der, which describes a distinct region living women to identify structures ative full bladder stress tests, had not had
of the body recognizable by most, the within the perineal body region, define previous surgery for pelvic floor disor-
words perineal body would cause most their 3-dimensional (3-D) location, and ders, and had Pelvic Organ Prolapse
to point to the region between the vagina provide a framework for visualizing this Quantification system (POP-Q) points
and anus. But the question remains, region’s complex anatomy. at least 1 cm above the hymenal ring.
what structures are involved in that re- Women were recruited by newspaper
gion? Is there a distinct, identifiable advertisements and matched for age,
structure called the perineal body, or is it M ATERIALS AND M ETHODS race, and parity with study subjects.
made up of its component pieces all tra- Thin-slice MRI scans were performed on Each woman underwent supine MRI
versing or inserting into this region? 11 women (all controls) in an ongoing at rest using a 3 Telsa Philips Achieva
scanner (Philips Medical Systems, Best,
From the Pelvic Floor Research Group and Division of Gynecology, Department of Obstetrics
The Netherlands) with a 6-channel
and Gynecology, University of Michigan, Ann Arbor, MI. phased array coil. Turbo spin echo tech-
Presented at the 30th Annual Meeting of the American Urogynecologic Society, Hollywood, FL,
niques were used to image the sagittal,
September 24-26, 2009. coronal, and axial planes. Thirty images
Received March 18, 2010; revised May 26, 2010; accepted June 7, 2010. were obtained in each plane (repetition
Reprints not available from the authors. time range, 2300 –3000, echo time 30, 0.2
This study was supported in part by the National Institute of Child Health and Human Development mm gap for 2 mm slices, and 1 mm gap
Grant R01 HD 38665 with additional investigator support from the Office for Research on for 5 mm slices, number of signal aver-
Women’s Health Specialized Centers of Research on Sex and Gender Factors Affecting Women’s ages 2, 256 ⫻ 255). To maximize visibil-
Health 1 P50 HD044406. J.O.L.D. receives research support from American Medical Systems
ity of small perineal body structures, 2
and is a consultant for Johnson and Johnson. D.E.F. receives research support from American
Medical Systems. The other authors have no disclosures to report. mm thick multiplanar proton density
0002-9378/$36.00 • © 2010 Mosby, Inc. All rights reserved. • doi: 10.1016/j.ajog.2010.06.008 MRIs were acquired in axial and coronal
planes.

NOVEMBER 2010 American Journal of Obstetrics & Gynecology 494.e15

Research Basic Science: Gynecology www.AJOG.org

FIGURE 1 TABLE 1
Building 3-D model Demographics
Study population
Characteristics (n ⴝ 11)
Age, ya 60 ⫾ 10
...........................................................................................................
BMI, kg/m 2a
24.8 ⫾ 4.7
...........................................................................................................
Median parity 2
...........................................................................................................
a
White 10 (91%)
...........................................................................................................
a
POP-Q
..................................................................................................
Ba –1.8 ⫾ 0.8
..................................................................................................
C –5.8 ⫾ 1.4
..................................................................................................
Bp –1.8 ⫾ 0.6
..................................................................................................
GH, rest 2.8 ⫾ 1.0
..................................................................................................
PB, rest 3.2 ⫾ 1.3
...........................................................................................................
a
Hysterectomy 0 (0%)
...........................................................................................................
Ba, Bartholins; BMI, body mass index; PB, perineal
body; POP-Q, Pelvic Organ Prolapse Quantification
system.
a
Mean ⫾ SD or n (%).
Larson. Perineal body anatomy in living women.
Am J Obstet Gynecol 2010.

Utilizing the description of the female

perineal body location by Oh and Kark,3
individual structures were identified by
the region bounded anteriorly by the
posterior vaginal wall and posteriorly by
the anterior anal wall on axial, sagittal,
and coronal scans. The longitudinal ex-
tent of the perineal body region was de-
fined as extending from the superficial
transverse perineal muscle and external
anal sphincter to the fusion of the longi-
tudinal muscle of the rectum with the in-
ternal anal sphincter, which occurs at the
caudal extent of the rectovaginal space.1
The structures were consistently iden-
tified by 2 different observers with over-
sight by the senior author. Images were
reviewed and a representative subject se-
lected for the clarity of her anatomy from
whom to generate a computer model of
this region using 3-D Slicer program
(version 2.1b1; Brigham and Women’s
Hospital, Boston, MA). The original 2 mm
axial and coronal Digital Imaging and
Communications in Medicine images
The top panel shows an axial image at the level of the vestibular bulb (VB). The middle panel shows were aligned, ensuring that structures co-
the outlines of the structures including pelvic bones (white), VB (turquoise), bulbosponsiosus (BS, localized in these axes by simultaneous re-
green), superficial transverse perineal (SPT; blue), external anal sphincter (EAS) (red), internal anal view of scan planes in the viewer.
sphincter (IAS; pink), and rectum (R; yellow). The bottom panel shows a 3-D model on MRI. A 3-D model was made of the pelvic
Larson. Perineal body anatomy in living women. Am J Obstet Gynecol 2010. bones, bladder, urethra, vagina, rectum,
and perineal body structures by tracing

494.e16 American Journal of Obstetrics & Gynecology NOVEMBER 2010

www.AJOG.org
FIGURE 2
Axial MRI scans showing perineal body (PB) anatomy
Arcuate pubic ligament (APL) as reference slice. Negative numbers are caudal and positive numbers are cephalad to APL. Table 2 lists the PB component
abbreviations.
B, bladder; Ba, Bartholins; IAS, internal anal sphincter; IC, iliococcygeus; PB, perineal body; PR, puborectalis; PS, pubic symphysis; PuR, pubic rami; R, rectum; U, urethra; V, vagina; VB, vestibular bulb.
Larson. Perineal body anatomy in living women. Am J Obstet Gynecol 2010.
the structure outlines on axial images
and creating 3-D models from these out-
lines (Figure 1). Each structure was vali-
dated by overlaying the model with orig-
inal source images in orthogonal planes.
Although only 1 model was made for
demonstration purposes, observations
were based on all 11 subjects.
Names conforming to Terminologia
Anatomica8 were used except for the le-
vator ani muscles in which we have cho-
sen the term pubovisceral muscle rather
than pubococcygeal muscle to more ac-
curately reflect its insertions.9 As de-
scribed by Kearney et al,9 the compo-
nents of the pubovisceral muscle are the
same as those referenced in the Termino-
logia Anatomica8 for the pubococcygeal
muscle: the pubovaginalis, puboperinea-
lis, and puboanalis.
R ESULTS
The mean age of the 11 study partici-
pants was 61 ⫾ 10 years (SD), mean body
Basic Science: Gynecology
mass index (BMI) was 24.8 ⫾ 4.7 kg/m2,
median parity was 2, and 91% were
white. POP-Q points are shown along
with the demographics in Table 1. No
subject reported having a significant ob-
stetrical laceration with their deliveries.
No subject had undergone a hysterec-
tomy or pelvic organ prolapse surgery.
None had pelvic floor dysfunction symp-
toms as determined by responses to vali-
dated questionnaires (PFDI and PFIQ).
Perineal body structures are best visu-
alized in the axial plane, revealing 3 dis-
tinct regions: superficial, mid, and deep.
In the superficial portion at the level of
the vestibular bulb (VB), the bulbospon-
giosus (BS) inserts into the lateral mar-
gins of the perineal body, whereas the
superficial transverse perineal muscle
(STP) and external anal sphincter (EAS)
traverse the region (Figure 2, panels –1.8
to – 0.6).
In the perineal body’s midregion at the
proximal end of the superficial trans-
NOVEMBER 2010 American Journal of Obstetrics & Gynecology
Research
verse perineal muscle, the puboperinea-
lis muscle (1 component of the pubovis-
ceral muscle) inserts into the lateral
margins of the perineal body and in
some individuals can be seen to cross the
midline. This region also contains the
distal internal anal sphincter. The pu-
boanalis muscle is also visible as it inserts
in the intersphincteric groove between
internal and external anal sphincters
(Figure 2, panels 0.0 to ⫹0.4).
The puboanalis muscle and internal anal
sphincter extend into the perineal body’s
most deep region at the level of the mi-
durethra. Here the pubovaginalis muscle
also becomes visible as it fuses with the vag-
inal side wall, sending fibers posteriorly to
the perineal body. In this location, the lon-
gitudinal muscle of the rectum may be vis-
ible in the midline (Figure 2, starting at
panel ⫹0.2 through ⫹0.8). The puborec-
talis muscle forms a loop behind the rec-
tum at this level but does not contribute
fibers to the perineal body.
494.e17
Research Basic Science: Gynecology
FIGURE 3
Sagittal MRI scans showing perineal body (PB) anatomy
Midsagittal slice as reference slice with “R” to the right. The PB is outlined in turquoise. Table 2 lists
the PB component abbreviations. The region that corresponds to the images shown in Figure 2 is
indicated; however, please note that this is a different subject from the one shown in Figure 2 so that
minor variations in spatial relationships exist.
B, bladder; IAS, internal anal sphincter; IC, iliococcygeus; PB, perineal body; PR, puborectalis; PS, pubic symphysis; PuR, pubic rami; R,
rectum; U, urethra; V, vagina; VB, vestibular bulb.
Larson. Perineal body anatomy in living women. Am J Obstet Gynecol 2010.
Although the perineal body’s struc-
tures are best visualized in the axial
plane, craniocaudal relationships can be
appreciated on sagittal images. In the
midsagittal plane, the perineal body’s
recognizable pyramidal structure is visi-
ble between the vagina and the rectum
(Figure 3, panel 0.0). As you move later-
ally, the perineal membrane is visible
above the bulbospongiosus muscle and
vestibular bulb in the midregion of the
perineal body. The fibers of the pubovis-
ceral muscle extend across the perineal
body, whereas the dorsal portion of the
external anal sphincter extends caudal to
the perineal body (Figure 3, panel R1.0).
These findings as well as the findings
on axial images are summarized in Table
2. Even though coronal images can de-
494.e18 American Journal of Obstetrics & Gynecology NOVEMBER 2010
lineate the immediate perianal struc-
tures, they are not as helpful for delineat-
ing the perineal body region because of
the orientation of the anatomy.
Figure 4 illustrates a 3-D model of the
perineal body generated from MRI
scans, further clarifying structural inter-
relationships and facilitating recognition
of the 3 perineal body regions: superfi-
cial, mid, and deep. The model helps
provide a framework for understanding
normal anatomy (Figure 4).
C OMMENT
This study describes the complex anat-
omy of the perineal body as seen in thin-
slice MRI of living women with normal
support and no pelvic floor dysfunction.
Axial images provide the clearest view of
www.AJOG.org
individual structures within this region,
whereas sagittal images illustrate cranio-
caudal relationships. This study suggests
a novel framework for organization of
this region into 3 parts: superficial, mid,
and deep. The findings we have de-
scribed were seen in the scans of all 11
subjects.
Our findings confirm published data
from cadaver and histological studies re-
garding the complexity of the perineal
body. Oh and Kark3 described this re-
gion between the “anterior anal wall and
the posterior vaginal wall” as a “heavily
used traffic center.” They acknowledge
the difficulty with dissecting this region
and, because of this, supplemented their
dissections with histological analysis.
Although their article focuses on the
male anatomy, in the female, they de-
scribe a region bounded by the vagina
anteriorly, the anterior anal wall posteri-
orly, and the transverse perinea muscles
and external anal sphincter muscles su-
perficially and discuss the difficulties in
identifying the proximal longitudinal
border. Although the authors describe it
as a solid, inseparable musculotendinous
mass, our MRI scans complement their
careful work and not only allow visual-
ization of distinct structures within this
region but also introduce the possibility
of generating a 3-D model.
Axial images allow visualization of the
pubovisceral portion of the levator mus-
cle in the perineal region (puboanalis,
puboperinealis, and pubovaginalis com-
ponents), which likely correlate with
what Oh and Kark3 describe as the
“pubopre-rectal” muscle. Similarly, we
show how the puborectalis muscle loops
behind the rectum likely helping sus-
pend the perineal body from the pubic
bone as described by Oh and Kark;3
however, it lies lateral to the iliococcy-
geal muscle and does not contribute fi-
bers to the perineal body.
Findings of the present study contra-
dict and extend what is currently in the
literature by utilizing advances in MRI
technology to better visualize this region
and describe its anatomy. In their MRI
study of the anal canal and perianal
structures, Morren et al2 describe the fe-
male perineal body as “clearly visible as a
separate anatomical structure located
www.AJOG.org Basic Science: Gynecology Research

TABLE 2
Characteristic magnetic resonance features of the perineal body (PB)
Structure Axial (Figure 3) Sagittal (Figure 4)
Superficial
.......................................................................................................................................................................................................................................................................................................................................................................
Superficial transverse perineal (STP) Extends laterally across midline, adjacent Not well seen
to EAS (–1.2 to –0.6); portion visible on
–1.8
.......................................................................................................................................................................................................................................................................................................................................................................
Bulbospongiosus (BS) Surrounds vestibular bulb on lateral Caudal to VB with dorsal fibers to PB (R1.0)
border (–1.2 to –0.6); portion visible on
–1.8
.......................................................................................................................................................................................................................................................................................................................................................................
External anal sphincter (EAS) Crosses midline ventrally, dorsal tear- Dorsal portion extends caudal to PB,
drop shape (–1.2 to ⫹0.4); portion ventral portion not well seen (R0.5 to R1.0)
visible on –1.8
................................................................................................................................................................................................................................................................................................................................................................................
Mid
.......................................................................................................................................................................................................................................................................................................................................................................
Internal anal sphincter (IAS) Innermost circular muscular layer (–1.2 Not well seen
to ⫹0.8)
.......................................................................................................................................................................................................................................................................................................................................................................
Pubovisceral portion of levator (PVi) Origins on pubic rami difficult to visualize PVi lateral to midsagittal extending across
Puboperinealis muscle (PP) in these cuts; however, can appreciate PB (R1.0); PA more lateral than PP/PVa
Puboanalis muscle (PA) fusion with appropriate viscera (0.0 to (R1.5)
⫹0.8)
.......................................................................................................................................................................................................................................................................................................................................................................
Perineal membrane (PM) Not well seen Cranial to VB, inserting into PB (R1.0)
.......................................................................................................................................................................................................................................................................................................................................................................
Still present: EAS
................................................................................................................................................................................................................................................................................................................................................................................
Deep
.......................................................................................................................................................................................................................................................................................................................................................................
Longitudinal fibers of rectum (LR) Ventral to IAS (⫹0.4 to ⫹0.8) Not well seen
.......................................................................................................................................................................................................................................................................................................................................................................
Pubovisceral portion of levator (PVi) PVa is ventral component of PVi that PVi lateral to midsagittal extending across
Pubovaginalis muscle (PVa) inserts into vagina (⫹0.4 to ⫹0.8) PB (R1.0); PA more lateral than PP/PVa
(R1.5)
.......................................................................................................................................................................................................................................................................................................................................................................
Still present: IAS, PA
................................................................................................................................................................................................................................................................................................................................................................................
Larson. Perineal body anatomy in living women. Am J Obstet Gynecol 2010.

between the posterior vaginal wall and including the EAS, a tendinous exten- women who were asymptomatic vagi-
anterior anal canal.” In addition to the sion of STP muscles, and a third layer of nally parous women in whom clinical
distinct perineal body, the authors de- tendinous fibers of the deep transverse testing confirmed normal support and
scribe the visible superficial transverse perineal muscles. In their cadaveric stud- questionnaires confirmed absence of
perineal muscle as well as a puboperinea- ies, a relationship between the levator ani pelvic floor dysfunction and known ob-
lis muscle, which likely corresponds to and the perineal body could not be stetrical laceration. This group repre-
our pubovisceral components. It is likely confirmed.6 sents the majority of women in the
that the stronger magnetic resonance This approach is different from our ap- reproductive age seen by clinicians.
magnet (3 Tesla vs 1.5 Tesla) and thinner proach to organize the anatomy of the per- Whether this anatomy is different in nul-
slices (2 mm vs 3mm) improved visibil- ineal body because it excludes several com- liparous women deserves further study.
ity in this region to allow determination ponents of this region visible on our The second consideration concerns the
of individual components within what advanced MRI and does include the deep impact of missing information not only
Morren et al2 recognized as a single transverse perineal muscles, which are not within the thickness of the MRI slices but
structure. evident in our study.6 In previous publica- also the spaces between each slice. We have
In addition, findings of the present tions, DeLancey1 discussed how this mus- tried to minimize this while studying this
study extend what is present in current cle has now been revised to include the region’s small structures by reducing the
literature by providing a different ap- compressor urethra and urethrovaginal slice thickness from the standard 5 mm to 2
proach to organizing the anatomy of this sphincter, which were not evident as part mm. Even using these thin slices, the ter-
region. In their published cadaveric and of that region in our images. mination fibers and connective tissue that
histological studies, Shafik et al6 pro- Several factors must be considered insert into bone are not visible on MRI. As
posed an approach to dividing the peri- when interpreting the results of this a result, our models incorporate the belly
neal body into 3 layers: a superficial layer study. This study was performed on of the muscles but not these insertions. Al-

NOVEMBER 2010 American Journal of Obstetrics & Gynecology 494.e19

Research Basic Science: Gynecology www.AJOG.org

FIGURE 4
3-D model of perineal body (PB)

A, Dorsal lithotomy view. B, Left lateral aspect of same image. C, With the removal of pelvic bones, a better appreciation of the superficial region with BS,
STP, and EAS is shown. D, Lateral view with BS removed to illustrate midregion at proximal STP with PP, PM, and PA. E, With removal of the EAS, STP,
PP, and PM the deep region with LR, PVa, and PA can be visualized. The IAS is visible with the EAS removed. The PVa muscle is barely visible lateral to
the vagina in this image.
B, bladder; G, gastrointestinal tract; IAS, internal anal sphincter; LR, longitudinal fibers of rectum; PA, puboanalis muscle; PM, perineal membrane; PP, puboperinealis muscle; PR, puborectalis; Pva,
pubovaginalis muscle; U, urethra; V, vagina.
Larson. Perineal body anatomy in living women. Am J Obstet Gynecol 2010.

though this needs to be kept in mind while
viewing the 3-D model, it does not have an
impact on the identification of the muscle
mass in the perineal body region.
Additionally, as described in the Re-
sults section, the identification of the
longitudinal muscle is limited by resolu-
tion of this magnetic resonance technol-
ogy. Endoanal MRI does capture this
anal region better because of proximity
to the coil.7,10
Lastly, this approach to studying anat-
omy with MRI and not histology from ca-
daveric dissections may seem counterin-
tuitive. However, we would argue that our
findings complement the cadaveric and
histological studies by Oh and Kark3 and
offer the benefit that utilizing images of liv-
ing women minimizes distortion that oc-
curs in cadaveric specimens. In the future,
this technique also allows studies to be
conducted on women with and without
known pelvic organ support defects,
something not possible in cadavers.
In summary, this study illustrates how
technological advances in MRI allow vi-
sualization of perineal body anatomy in
living women without pelvic floor dys-
function and help to unravel the mystery
behind its components. Organization of
this region into 3 parts, superficial, mid,
and deep, provides a framework for vi-
sualizing its normal anatomy. MRI-gen-

494.e20 American Journal of Obstetrics & Gynecology NOVEMBER 2010

www.AJOG.org
erated 3-D models greatly expand our
ability to understand the structural rela-
tionships of the complex yet frequently
encountered region. Just as MRI is used
to direct surgical repairs in the shoulder
by helping us gain insight into normal
anatomy, MRI of the perineal body
moves us 1 step closer to identifying spe-
cific failures and appropriately tailoring
surgical repairs in this region of the body
as well. f Physioanatomical relationship of the external
REFERENCES
1. DeLancey JO. Structural anatomy of the pos-
terior pelvic compartment as it relates to recto-
cele. Am J Obstet Gynecol 1999;180:815-23.
Basic Science: Gynecology
2. Morren GL, Beets-Tan RG, van Engelshoven
JM. Anatomy of the anal canal and perianal
structures as defined by phased-array mag-
netic resonance imaging. Br J Surg 2001;88:
1506-12.
3. Oh C, Kark AE. Anatomy of the perineal
body. Dis Colon Rectum 1973;16:444-54.
4. Shafik A, Ahmed I, Shafik AA, El-Ghamrawy
TA, El-Sibai O. Surgical anatomy of the perineal
muscles and their role in perineal disorders.
Anat Sci Int 2005;80:167-71.
5. Shafik A, Shafik IA, el-Sibai O, Shafik AA.
anal sphincter to the bulbocavernosus muscle
in the female. Int Urogynecol J Pelvic Floor Dys-
funct 2007;18:851-6.
6. Shafik A, Sibai OE, Shafik AA, Shafik IA. A
novel concept for the surgical anatomy of the
NOVEMBER 2010 American Journal of Obstetrics & Gynecology
Research
perineal body. Dis Colon Rectum 2007;50:
2120-5.
7. Stoker J, Wallner C. The anatomy of the pel-
vic floor and sphincters. In: Stoker J, Taylor SA,
DeLancey JO, eds. Imaging pelvic floor disor-
ders, 2nd ed. Heidelberg, Germany: Springer-
Verlag; 2008:19-29.
8. Federative Committee on Anatomical Termi-
nology. Terminologia Anatomica. New York:
Thieme Stuttgart, Germany; 1998.
9. Kearney R, Sawhney R, DeLancey JO. Levator
ani muscle anatomy evaluated by origin-insertion
pairs. Obstet Gynecol 2004;104:168-73.
10. Hussain SM, Stoker J, Lameris JS. Anal
sphincter complex: endoanal MR imaging of
normal anatomy. Radiology 1995;197:671-7.
494.e21