Abstract
Background: Adenomyosis is a benign uterine disorder that commonly presents with heavy menstrual
bleeding, dysmenorrhoea, chronic pelvic pain and infertility. Although histopathological examination of
hysterectomy specimens remains the reference standard, non-invasive imaging has become increasingly
important for preoperative diagnosis. Transvaginal ultrasonography is widely used as a first-line
investigation, whereas magnetic resonance imaging offers superior soft-tissue contrast and detailed
evaluation of the uterine junctional zone.
Aim: To compare the diagnostic performance of transvaginal ultrasonography and magnetic resonance
imaging in the evaluation of adenomyosis.
Materials and methods
This hos
pital-based comparative cross-sectional diagnostic-accuracy study
included 120 women with clinical suspicion of adenomyosis who underwent hysterectomy at a tertiary-care
hospital. All participants were evaluated preoperatively using TVUS and pelvic MRI. Imaging findings
were compared with histopathological examination of hysterectomy specimens as the reference standard.
Sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratios, diagnostic
accuracy and diagnostic odds ratios were calculated with 95% confidence intervals. Agreement was
assessed using Cohen’s kappa coefficient, and the paired diagnostic performance of TVUS and MRI was
compared using the exact McNemar test. A p value <0.05 was considered statistically significant.
Results
Histopathological examination confirmed adenomyosis in 68 (56.7%) of the 120 participants.
TVUS correctly identified 57 true-positive and 44 true-negative cases, whereas MRI identified 64 true-
positive and 49 true-negative cases. TVUS demonstrated sensitivity of 83.8%, specificity of 84.6%,
positive predictive value of 87.7%, negative predictive value of 80.0% and diagnostic accuracy of 84.2%.
The corresponding values for MRI were 94.1%, 94.2%, 95.5%, 92.5% and 94.2%, respectively. MRI had a
higher positive likelihood ratio (16.31 versus 5.45), a lower negative likelihood ratio (0.06 versus 0.19) and
a higher diagnostic odds ratio (261.33 versus 28.50). MRI showed almost-perfect agreement with
histopathology (κ=0.88; p<0.001), whereas TVUS showed substantial agreement (κ=0.68; p<0.001). MRI
demonstrated significantly higher sensitivity (p=0.039), lower false-negative rate (5.9% versus 16.2%;
p=0.039) and greater overall diagnostic accuracy than TVUS (94.2% versus 84.2%; p=0.004).
Conclusion
TVUS provided good diagnostic performance and remained a suitable first-line imaging
modality for adenomyosis. MRI demonstrated higher sensitivity, accuracy and agreement with
histopathology and produced fewer false-negative findings. MRI should be considered when TVUS
findings are inconclusive or when detailed assessment of the extent of adenomyosis is required.
Keywords
Adenomyosis; Transvaginal ultrasonography; Magnetic resonance imaging.
Introduction
Adenomyosis is a benign uterine disorder characterised by the presence of endometrial glands
and stroma within the myometrium, accompanied by hypertrophy and hyperplasia of the
surrounding smooth muscle. It commonly affects women of reproductive and perimenopausal
age and may present with heavy menstrual bleeding, progressive dysmenorrhoea, chronic pelvic
pain, dyspareunia, infertility or an enlarged tender uterus. However, some patients remain
asymptomatic. The clinical presentation frequently overlaps with uterine leiomyomas and
endometriosis, making diagnosis based only on symptoms and pelvic examination difficult.
Historically, adenomyosis was confirmed by histopathological examination of hysterectomy
specimens. Increasing use of non-invasive imaging has facilitated preoperative diagnosis and the
conservative management of women wishing to preserve fertility or avoid hysterectomy [1,2].
Transvaginal ultrasonography (TVUS) is generally used as the first-line imaging modality
because it is widely available, relatively inexpensive, well tolerated and capable of providing
~ 353 ~
real-time assessment of the uterus. Sonographic features
suggestive of adenomyosis include globular uterine enlargement,
asymmetrical myometrial thickening, myometrial cysts,
hyperechoic islands, fan -shaped shadowing, echogenic
subendometrial lines and buds, translesional vascularity and an
irregular or interrupted junctional zone. The Morphological
Uterus Sonographic Assessment consensus has improved the
terminology and standardisation of these findings [3,4].
Nevertheless, the diagno stic performance of TVUS remains
dependent on operator experience, equipment quality and the
presence of coexisting pelvic abnormalities.
Magnetic resonance imaging (MRI) provides superior soft-tissue
contrast, multiplanar assessment and improved visualisa tion of
the uterine junctional zone. MRI findings associated with
adenomyosis include junctional -zone thickening, an increased
junctional-zone-to-myometrial thickness ratio, ill -defined areas
of low signal intensity and small high -signal-intensity
myometrial foci on T2 - or T1 -weighted images. MRI may be
particularly useful when TVUS findings are inconclusive or
when adenomyosis coexists with leiomyomas or endometriosis.
However, MRI is more expensive, less accessible and
contraindicated in certain patients. Available evidence suggests
that both TVUS and MRI have good diagnostic performance,
although results vary according to diagnostic criteria, examiner
expertise and the reference standard employed [2,5]. A direct
comparison of these modalities against hist opathological
findings may help define their relative diagnostic value and
guide appropriate imaging choices. Therefore, the present study
compared TVUS and MRI in the evaluation of adenomyosis
among women undergoing hysterectomy.
Aim
To compare the diag nostic performance of transvaginal
ultrasonography and magnetic resonance imaging in the
evaluation of adenomyosis.
Objectives
1. To determine the sensitivity, specificity, positive predictive
value, negative predictive value and diagnostic accuracy of
transvaginal ultrasonography in detecting adenomyosis.
2. To determine the sensitivity, specificity, positive predictive
value, negative predictive value and diagnostic accuracy of
magnetic resonance imaging in detecting adenomyosis.
3. To compare the findings of t ransvaginal ultrasonography
and magnetic resonance imaging with histopathological
examination and assess their diagnostic agreement.
Material
and Methodology
Source of Data
The study participants were recruited from women attending the
outpatient and inpatient services of the Department of Obstetrics
and Gynaecology at a tertiary -care teaching hospital. Women
who had symptoms or clinical findings suggestive of
adenomyosis and who were scheduled to undergo hysterectomy
were assessed for eligibility. Clinic al information, TVUS
findings, MRI findings, operative observations and
histopathological results were obtained from clinical
examination records, radiological reports, operative notes and
histopathology reports.
Study Design
A hospital -based comparative c ross-sectional diagnostic -
accuracy study was conducted. Both TVUS and MRI were
performed in the same participants, and the results of each
imaging modality were compared with histopathological
examination of hysterectomy specimens, which was considered
the reference standard.
Study Location
The study was conducted jointly in the Departments of
Obstetrics and Gynaecology, Radiodiagnosis and Pathology.
Study Duration
The study was conducted over 12 months. This period included
participant recruitment, imagin g assessment, surgical
management, histopathological examination, data collection and
statistical analysis.
Sample Size
A total of 120 eligible women were included in the study. The
sample size was estimated using the anticipated sensitivity of
imaging fo r diagnosing adenomyosis, an expected prevalence
among symptomatic women undergoing hysterectomy, a 95%
confidence level and an acceptable absolute precision. It was
calculated using the diagnostic-study formula:
where
at a 95% confidence level,
represented
the anticipated sensitivity,
represented absolute precision and
represented the anticipated prevalence of adenomyosis. After
allowing for incomplete investigations or records, the final
sample size was fixed at 120 participants. Consecutive eligible
participants were recruited until the required sample size was
achieved.
Inclusion Criteria
Women were included when they:
Were aged 18 years or older.
Presented with symptoms or clinical findings sugges tive of
adenomyosis, such as heavy menstrual bleeding,
dysmenorrhoea, chronic pelvic pain, dyspareunia or a bulky
tender uterus.
Were scheduled to undergo hysterectomy as part of their
clinical management.
Underwent both TVUS and pelvic MRI before surgery.
Provided written informed consent to participate in the
study.
Had histopathological examination of the hysterectomy
specimen available for comparison.
Exclusion Criteria
Women were excluded when they:
Were pregnant or suspected to be pregnant.
Had contraindications to MRI, including non -compatible
metallic implants, cardiac pacemakers or severe
claustrophobia.
Were unwilling or unable to undergo TVUS.
Had previously undergone endometrial ablation, uterine
artery embolisation or conservative surge ry for
adenomyosis.
Had received hormonal treatment that could substantially
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 354 ~
alter imaging findings during the preceding three months.
Had a known or suspected uterine, cervical or ovarian
malignancy.
Had technically inadequate TVUS or MRI examinations.
Did not undergo hysterectomy or had an
unavailable/inadequate specimen for histopathological
examination.
Declined participation or had incomplete essential study
data.
Procedure and Methodology
Approval was obtained from the Institutional Ethics Comm ittee
before commencing the study. Written informed consent was
obtained from every participant. Confidentiality of patient
information was maintained throughout the study.
A detailed history was recorded, including age, parity, menstrual
pattern, heavy me nstrual bleeding, severity and duration of
dysmenorrhoea, chronic pelvic pain, dyspareunia, infertility,
previous caesarean delivery, uterine surgery and hormonal
treatment. General, abdominal and pelvic examinations were
performed. Relevant laboratory findings, including haemoglobin
concentration, were documented.
Transvaginal ultrasonography
TVUS was performed using a high -frequency endovaginal
transducer, preferably during the early proliferative phase when
clinically feasible. The urinary bladder was e mptied before
examination. The uterus was assessed in sagittal and transverse
planes for its position, dimensions, volume, shape, symmetry of
the uterine walls, myometrial echotexture and associated pelvic
abnormalities.
Adenomyosis was suspected when one or more recognised
MUSA features were identified, including:
Myometrial cysts;
Hyperechoic islands;
Echogenic subendometrial lines or buds;
Fan-shaped acoustic shadowing;
Globular uterine configuration;
Asymmetrical myometrial thickening;
Translesional vascularity;
An irregular or interrupted junctional zone.
Direct and indirect features were recorded separately. The
location and extent of suspected adenomyosis, involvement of
the inner or outer myometrium and the presence of focal or
diffuse disea se were documented. Associated leiomyomas,
endometriosis or adnexal pathology were also recorded. The
TVUS examiner recorded the final impression as adenomyosis
present or absent.
Magnetic resonance imaging
Pelvic MRI was performed using a 1.5 -Tesla or 3 -Tesla MRI
system with a phased -array pelvic coil. The standard protocol
included multiplanar T2 -weighted sequences and axial T1 -
weighted sequences, with fat -suppressed sequences wherever
required. Contrast-enhanced imaging was performed only when
clinically indicated.
MRI evaluation included uterine size and configuration,
myometrial signal intensity, maximum junctional -zone
thickness, junctional -zone irregularity, the ratio of maximum
junctional-zone thickness to total myometrial thickness and the
presence of small high -signal-intensity myometrial foci.
Adenomyosis was diagnosed when characteristic features such
as junctional-zone thickening, an ill-defined low-signal-intensity
myometrial area or punctate high -signal-intensity foci were
identified. The dise ase was classified as focal or diffuse and its
location and extent were recorded.
Whenever feasible, the radiologist interpreting the MRI was
blinded to the detailed TVUS interpretation, and the imaging
specialists were blinded to the final histopathologic al findings.
The interval between imaging and surgery was kept as short as
practically possible, preferably within four weeks.
Following imaging, the participants underwent hysterectomy
according to their clinical indications and institutional treatment
protocols. The decision to perform surgery was made by the
treating clinician and was independent of participation in the
study.
Sample Processing
Immediately after hysterectomy, the uterus was labelled with a
unique identification number and transported to the Department
of Pathology in 10% neutral -buffered formalin. Gross
examination included uterine dimensions, external appearance,
endometrial cavity, myometrial thickness, areas of haemorrhage,
cystic changes, trabeculation and coexisting leiomyomas or
other lesions.
Representative tissue sections were obtained from the anterior
and posterior uterine walls, fundus, cornual regions and any
grossly suspicious myometrial area. The tissue was processed
routinely, embedded in paraffin, sectioned at approximatel y 3-5
µm and stained with haematoxylin and eosin.
Adenomyosis was diagnosed histopathologically by the presence
of endometrial glands and stroma within the myometrium,
accompanied by surrounding smooth -muscle hypertrophy or
hyperplasia. The pathologist was blinded, as far as feasible, to
the final imaging diagnosis. Histopathological findings were
categorised as adenomyosis present or absent and were used as
the reference standard.
Statistical Methods
Data were entered into Microsoft Excel and analysed usi ng IBM
SPSS Statistics or an equivalent statistical package. Continuous
variables were assessed for normality using the Shapiro -Wilk
test and graphical methods. Normally distributed variables were
summarised as mean and standard deviation with 95%
confidence intervals, whereas skewed variables were reported as
median and interquartile range. Categorical variables were
presented as frequencies and percentages.
The diagnostic performances of TVUS and MRI were calculated
against histopathology in terms of:
Sensitivity;
Specificity;
Positive predictive value;
Negative predictive value;
False-positive and false-negative rates;
Overall diagnostic accuracy;
Positive and negative likelihood ratios.
All diagnostic estimates were reported with 95% confidence
intervals. The chi -square test or Fisher’s exact test was used to
assess associations between categorical variables. Because
TVUS and MRI results were obtained from the same
participants, their paired proportions, including sensitivity,
specificity and over all accuracy, were compared using
McNemar’s test. Agreement between each imaging modality and
histopathology was assessed using Cohen’s kappa coefficient.
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 355 ~
Receiver operating characteristic curves were constructed for
relevant continuous imaging measurement s, such as junctional -
zone thickness. The area under the curve and its 95% confidence
interval were calculated, and an appropriate diagnostic cut -off
was identified using the Youden index. A two -sided p value
<0.05 was considered statistically significant.
Data Collection
Data were collected using a predesigned, pretested and
structured case -record form. The form contained the following
sections:
Sociodemographic details;
Obstetric and menstrual history;
Presenting symptoms and their duration;
Previous medical, hormonal and surgical history;
General and pelvic examination findings;
TVUS characteristics and diagnostic impression;
MRI characteristics and diagnostic impression;
Operative details;
Gross pathological findings;
Final histopathological diagnosis.
Each participant was assigned a unique study code. The
completed forms were checked regularly for accuracy,
consistency and completeness. Data were entered into a
password-protected electronic database, and participant
identifiers were removed before statistical analysis.
Observation and Results
Table 1: Comparison of the diagnostic performance of TVUS and MRI for adenomyosis (N=120)
Diagnostic parameter TVUS, n/N (%) 95% CI MRI, n/N (%) 95% CI Test of significance p value
Sensitivity 57/68 (83.8) 73.3-90.7 64/68 (94.1) 85.8-97.7 Exact McNemar test 0.039
Specificity 44/52 (84.6) 72.5-92.0 49/52 (94.2) 84.4-98.0 Exact McNemar test 0.125
Positive predictive value 57/65 (87.7) 77.5-93.6 64/67 (95.5) 87.6-98.5 Difference=7.8% 0.103
Negative predictive value 44/55 (80.0) 67.6-88.4 49/53 (92.5) 82.1-97.0 Difference=12.5% 0.061
Diagnostic accuracy 101/120 (84.2) 76.6-89.6 113/120 (94.2) 88.4-97.1 Exact McNemar test 0.004
False-positive rate 8/52 (15.4) 8.0-27.5 3/52 (5.8) 2.0-15.6 Exact McNemar test 0.125
False-negative rate 11/68 (16.2) 9.3-26.7 4/68 (5.9) 2.3-14.2 Exact McNemar test 0.039
Positive likelihood ratio 5.45 2.88-10.31 16.31 5.44-48.88
Negative likelihood ratio 0.19 0.11-0.34 0.06 0.02-0.15
Diagnostic odds ratio 28.50 10.57-76.86 261.33 55.88-1222.08
Table 1 compares the diagnostic performance of TVUS and MRI
for detecting adenomyosis among 120 participants. MRI
demonstrated significantly greater sensitivity than TVUS (94.1%
versus 83.8%; p=0.039). Although MRI also showed higher
specificity (94.2% versus 84.6%), the difference was not
statistically significant (p=0.125). The positive predictive value
was 95.5% for MRI and 87.7% for TVUS (p=0.103), while the
corresponding negative predictive values were 92.5% and 80.0%
(p=0.061). MRI achieved significantly higher overall diagnostic
accuracy than TVUS (94.2% versus 84.2%; p=0.004). MRI also
had lower false -positive (5.8% versus 15.4%; p=0.125) and
significantly lower false -negative rates (5.9% versus 16.2%;
p=0.039). F urthermore, MRI had a higher positive likelihood
ratio (16.31 versus 5.45), a lower negative likelihood ratio (0.06
versus 0.19) and a substantially higher diagnostic odds ratio
(261.33 versus 28.50), indicating superior diagnostic
performance.
A. TVUS-histopathology cross-classification
Table 2: Diagnostic performance of TVUS for detecting adenomyosis using histopathology as the reference standard (N=120)
TVUS finding Histopathology positive, n (%) Histopathology negative, n (%) Total, n (%)
Positive for adenomyosis 57 (47.5) True positive 8 (6.7) False positive 65 (54.2)
Negative for adenomyosis 11 (9.2) False negative 44 (36.7) True negative 55 (45.8)
Total 68 (56.7) 52 (43.3) 120 (100.0)
Association between TVUS and histopathological diagnosis: χ²=55.59, p<0.001.
B. Diagnostic indices of TVUS
Diagnostic index Estimate 95% CI Test statistic p value
Sensitivity 83.8% 73.3-90.7
Specificity 84.6% 72.5-92.0
Positive predictive value 87.7% 77.5-93.6
Negative predictive value 80.0% 67.6-88.4
Diagnostic accuracy 84.2% 76.6-89.6
Positive likelihood ratio 5.45 2.88-10.31
Negative likelihood ratio 0.19 0.11-0.34
Diagnostic odds ratio 28.50 10.57-76.86 χ²=55.59 <0.001
Agreement with histopathology κ=0.68 0.55-0.81 Z=7.47 <0.001
Table 2 presents the diagnostic performance of TVUS using
histopathology as the reference standard. Histopathological
examination confirmed adenomyosis in 68 (56.7%) participants
and excluded it in 52 (43.3%). TVUS correctly identified 57
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 356 ~
true-positive cases and 44 true-negative cases, while producing 8
false-positive and 11 false -negative results. TVUS had a
sensitivity of 83.8% (95% CI: 73.3 -90.7), specificity of 84.6%
(95% CI: 72.5 -92.0), positive predictive value of 87.7% (95%
CI: 77.5 -93.6), negative predictive value of 80.0% (95% CI:
67.6-88.4) and overall diagnostic accuracy of 84.2% (95% CI:
76.6-89.6). Its positive and negative likelihood ratios were 5.45
and 0.19, respectively. The diagnostic odds ratio was 28.50
(95% CI: 10.57 -76.86), and the association between TVUS and
histopathological diagnosis was statistically significant
(χ²=55.59, p<0.001). TVUS also demonstrated substantial
agreement with histopathology (κ=0.68; 95% CI: 0.55 -0.81;
Z=7.47; p<0.001).
A. MRI-histopathology cross-classification
Table 3: Diagnostic performance of MRI for detecting adenomyosis using histopathology as the reference standard (N=120)
MRI finding Histopathology positive, n (%) Histopathology negative, n (%) Total, n (%)
Positive for adenomyosis 64 (53.3) True positive 3 (2.5) False positive 67 (55.8)
Negative for adenomyosis 4 (3.3) False negative 49 (40.8) True negative 53 (44.2)
Total 68 (56.7) 52 (43.3) 120 (100.0)
Association between MRI and histopathological diagnosis: χ²=93.27, p<0.001.
B. Diagnostic indices of MRI
Diagnostic index Estimate 95% CI Test statistic p value
Sensitivity 94.1% 85.8-97.7
Specificity 94.2% 84.4-98.0
Positive predictive value 95.5% 87.6-98.5
Negative predictive value 92.5% 82.1-97.0
Diagnostic accuracy 94.2% 88.4-97.1
Positive likelihood ratio 16.31 5.44-48.88
Negative likelihood ratio 0.06 0.02-0.15
Diagnostic odds ratio 261.33 55.88-1222.08 χ²=93.27 <0.001
Agreement with histopathology κ=0.88 0.78-0.98 Z=9.66 <0.001
Table 3 shows the diagnostic performance of MRI using
histopathological examination as the reference standard. MRI
correctly identified 64 of the 68 histopathologically positive
cases and 49 of the 52 histopathologically negative cases. Only 3
false-positive and 4 false -negative results were observed. MRI
demonstrated a sensitivity of 94.1% (95% CI: 85.8 -97.7),
specificity of 94.2% (95% CI: 84.4 -98.0), positive predictive
value of 95.5% (95% CI: 87.6 -98.5), negative predictive value
of 92.5% (95% CI: 82.1-97.0) and diagnostic accuracy of 94.2%
(95% CI: 88.4 -97.1). The positive likelihood ratio was 16.31,
whereas the negative likelihood ratio was 0.06. The diagnostic
odds ratio was 261.33 (95% CI: 55.88 -1222.08), and MRI
findings were significantly associated with histopathological
diagnosis (χ²=93.27, p<0.001). MRI demonstrated almost -
perfect agreement with histopathology (κ=0.88; 95% CI: 0.78 -
0.98; Z=9.66; p<0.001).
A. Agreement of each imaging modality with histopathology
Table 4: Agreement and paired comparison of TVUS and MRI findings with histopathology (N=120)
Comparison Observed agreement, n/N (%) 95% CI Cohen’s κ 95% CI for κ Strength of agreement Test statistic p value
TVUS versus histopathology 101/120 (84.2) 76.6-89.6 0.68 0.55-0.81 Good/substantial Z=7.47 <0.001
MRI versus histopathology 113/120 (94.2) 88.4-97.1 0.88 0.78-0.98 Almost perfect Z=9.66 <0.001
TVUS versus MRI diagnosis 104/120 (86.7) 79.4-91.6 0.73 0.61-0.85 Good/substantial Z=8.03 <0.001
B. Paired comparison of correctly classified cases
TVUS classification MRI correct, n (%) MRI incorrect, n (%) Total, n (%)
TVUS correct 99 (82.5) 2 (1.7) 101 (84.2)
TVUS incorrect 14 (11.7) 5 (4.2) 19 (15.8)
Total 113 (94.2) 7 (5.8) 120 (100.0)
Exact McNemar test comparing overall diagnostic accuracy: p=0.004.
C. Summary of paired diagnostic differences
Diagnostic comparison TVUS MRI Absolute difference Test p value
Correctly classified cases 101 (84.2%) 113 (94.2%) +10.0 percentage points Exact McNemar 0.004
Histopathology-positive cases detected 57/68 (83.8%) 64/68 (94.1%) +10.3 percentage points Exact McNemar 0.039
Histopathology-negative cases correctly excluded 44/52 (84.6%) 49/52 (94.2%) +9.6 percentage points Exact McNemar 0.125
False-negative cases 11/68 (16.2%) 4/68 (5.9%) −10.3 percentage points Exact McNemar 0.039
False-positive cases 8/52 (15.4%) 3/52 (5.8%) −9.6 percentage points Exact McNemar 0.125
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 357 ~
Table 4 demonstrates the agreement and paired comparison of
TVUS and MRI with histopathological findings. TVUS agreed
with histopathology in 101 of 120 cases, producing an observed
agreement of 84.2% (95% CI: 76.6 -89.6) and substantial
agreement beyond chance (κ=0.68; p<0.001). MRI agreed with
histopathology in 113 cas es, corresponding to an observed
agreement of 94.2% (95% CI: 88.4 -97.1) and almost -perfect
agreement (κ=0.88; p<0.001). Agreement between TVUS and
MRI diagnoses was 86.7% (κ=0.73; 95% CI: 0.61 -0.85;
p<0.001). Both modalities correctly classified 99 (82.5%) cases;
however, MRI correctly classified 14 cases that were incorrectly
classified by TVUS, whereas TVUS correctly classified only 2
cases that were incorrectly classified by MRI. Consequently,
MRI showed significantly higher overall accuracy than TVUS
(94.2% versus 84.2%; exact McNemar p=0.004). MRI detected
10.3 percentage points more histopathology -positive cases
(p=0.039) and reduced the false-negative rate by 10.3 percentage
points (p=0.039). Although MRI also correctly excluded 9.6
percentage points more disease-negative cases and reduced the
false-positive rate by 9.6 percentage points, these differences
were not statistically significant (p=0.125).
Discussion
In the present comparative study of 120 women,
histopathological examination confirmed ade nomyosis in 68
(56.7%) participants. Both transvaginal ultrasonography (TVUS)
and magnetic resonance imaging (MRI) demonstrated good
diagnostic performance; however, MRI consistently produced
higher sensitivity, specificity, predictive values, diagnostic
accuracy and agreement with histopathology. The improved
performance of MRI was particularly evident in its lower false -
negative rate and significantly higher overall diagnostic
accuracy. These findings support the use of TVUS as the initial
imaging investi gation and MRI as a valuable second -line or
problem-solving modality, especially when TVUS findings are
inconclusive or complex coexisting uterine pathology is present.
Comparison of TVUS and MRI diagnostic performance
The sensitivity of MRI in the presen t study was significantly
higher than that of TVUS (94.1% versus 83.8%; p=0.039). MRI
also showed higher specificity (94.2% versus 84.6%), although
this difference was not statistically significant (p=0.125).
Similarly, MRI demonstrated higher positive and negative
predictive values than TVUS. The overall diagnostic accuracy
was 94.2% for MRI compared with 84.2% for TVUS,
representing a significant absolute improvement of 10
percentage points (p=0.004).
Van den Bosch et al . (2015) [1] developed the Morpholo gical
Uterus Sonographic Assessment terminology to standardise the
reporting of myometrial and junctional -zone abnormalities.
Their consensus identified myometrial cysts, hyperechoic
islands, fan -shaped shadowing, asymmetrical myometrial
thickening, transl esional vascularity and junctional -zone
irregularity as important sonographic features of adenomyosis.
Consistent application of these criteria may explain the relatively
high TVUS sensitivity observed in the present study.
Pinzauti et al . (2015) [2] reported that TVUS performed by
experienced examiners could reliably identify adenomyosis
before surgery. However, they also emphasised that diagnostic
performance depended on examiner expertise and the presence
of coexisting uterine lesions. Bazot and Daraï (2018) [3] similarly
observed that TVUS and MRI were the principal non -invasive
diagnostic modalities, with MRI providing superior tissue
characterisation and more objective evaluation of the junctional
zone.
In a systematic review and meta-analysis, Andres et al. (2018) [4]
concluded that TVUS had good diagnostic performance and
could be used as a first -line investigation for adenomyosis.
However, variation in individual sonographic criteria and
operator experience contributed to heterogeneity across studies .
Liu et al . (2019) [5] compared commonly used diagnostic
approaches and found that both TVUS and MRI had clinically
useful accuracy, although MRI generally showed greater
specificity. This is consistent with the present finding of 94.2%
specificity for MRI compared with 84.6% for TVUS.
Tellum et al . (2020) [6] found comparable overall diagnostic
ability for TVUS and MRI in their meta -analysis, but MRI
showed an advantage in specificity and in diagnostically difficult
cases. Chapron et al . (2020) [7] recommended an integrated
clinical and imaging approach and observed that expert TVUS
could perform well, while MRI was helpful for lesion mapping,
disease classification and assessment of the junctional zone. The
present findings similarly demonstrated that bot h modalities
were diagnostically useful, although MRI achieved significantly
higher accuracy.
The higher MRI performance in this study could be attributed to
its superior soft-tissue contrast, multiplanar capability and ability
to demonstrate junctional -zone thickening, myometrial signal
abnormalities and small haemorrhagic or cystic foci. By
contrast, TVUS performance may be influenced by uterine
position, obesity, coexisting leiomyomas and operator
interpretation. Harmsen et al . (2022) [8] revised the MUS A
definitions by dividing sonographic signs into direct and indirect
features, thereby improving their clinical interpretation and
reproducibility. Application of these revised criteria may further
improve future TVUS performance.
Diagnostic performance of TVUS
TVUS correctly detected 57 of the 68 histopathologically
confirmed cases and correctly excluded disease in 44 of the 52
histopathology-negative participants. It produced 8 false-positive
and 11 false -negative findings. Its sensitivity was 83.8%,
specificity 84.6%, positive predictive value 87.7%, negative
predictive value 80.0% and diagnostic accuracy 84.2%. The
diagnostic odds ratio was 28.50, and TVUS demonstrated
substantial agreement with histopathology (κ=0.68, p<0.001).
The TVUS sensitivity obs erved in this study was higher than
some pooled estimates reported by Andres et al. (2018) [4] and
Tellum et al . (2020) [6]. This may reflect differences in study
populations, prevalence, sonographer expertise, equipment
quality and the number of sonograph ic criteria required for
diagnosis. Studies limited to women undergoing hysterectomy
may also contain a higher proportion of advanced or diffuse
adenomyosis, in which ultrasonographic features are more
apparent.
Vannuccini et al . (2017) [9] explained that adenomyosis may
occur in diffuse, focal or mixed forms and is frequently
associated with smooth -muscle hypertrophy and disruption of
the endometrial -myometrial junction. These variable
pathological patterns may explain false-negative TVUS findings,
particularly in early or focal disease. Abbott (2017) [10] also
noted that adenomyosis is a heterogeneous disorder and that
variations in diagnostic criteria contribute to inconsistencies
between imaging and histopathology.
Exacoustos et al. (2020) [11] demonstrated that the number and
type of ultrasound features were associated with the severity and
distribution of adenomyosis. Direct signs, such as myometrial
cysts and hyperechoic islands, were generally more specific,
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 358 ~
whereas indirect signs, such as a globular uterus and
asymmetrical wall thickening, could also occur with leiomyomas
or other myometrial disorders. This could explain the eight false-
positive TVUS findings in the present study.
Tellum et al. (2020) [12] reported that individual ultrasound signs
did not have uniform diagnostic value and that combining
several features improved diagnostic confidence. Rasmussen et
al. (2019) [13] similarly found that interobserver agreement
differed across ultrasonographic features, indicating that
examiner training and standardised reporting are essential. In the
present study, the positive likelihood ratio of 5.45 indicated that
a positive TVUS result produced a moderate increase in the
probability of adenomyosis, while the negative likelihood ratio
of 0.19 showed tha t a negative examination substantially
reduced but did not completely exclude the disease.
The substantial agreement between TVUS and histopathology
(κ=0.68) suggests that TVUS was clinically reliable but not
interchangeable with the reference standard. Va n den Bosch et
al. (2019) [14] highlighted that structured assessment of the
myometrium and junctional zone could improve diagnostic
reproducibility. Therefore, TVUS remains an appropriate first -
line modality because it is inexpensive, widely available and
provides real -time assessment without radiation exposure.
Nevertheless, negative TVUS findings should be interpreted
cautiously when clinical suspicion remains high.
Diagnostic performance of MRI
MRI correctly identified 64 of the 68 histopathologically
positive cases and correctly excluded adenomyosis in 49 of the
52 disease -negative participants. Only 3 false -positive and 4
false-negative findings were recorded. MRI demonstrated
sensitivity of 94.1%, specificity of 94.2%, positive predictive
value of 95 .5%, negative predictive value of 92.5% and overall
diagnostic accuracy of 94.2%. Its positive likelihood ratio was
16.31, negative likelihood ratio was 0.06 and diagnostic odds
ratio was 261.33. These findings indicated that MRI was highly
effective for both confirming and excluding adenomyosis.
Agostinho et al . (2017) [15] described MRI as a highly useful
modality for evaluating adenomyosis because of its ability to
identify junctional -zone thickening, ill -defined low -signal-
intensity areas and high -signal-intensity myometrial foci. The
authors also cautioned that junctional -zone thickness could be
influenced by age, menstrual phase, hormonal status and
transient uterine contractions. These factors could account for
occasional false-positive or false-negative MRI findings.
Bazot and Daraï (2018) [3] reported that MRI was especially
valuable when ultrasonographic findings were equivocal or
when adenomyosis coexisted with uterine leiomyomas. The high
MRI specificity in the present study supports this observat ion.
Tellum et al. (2020) [6] also found that MRI generally provided
higher specificity than TVUS, although pooled sensitivity
estimates varied across studies.
Moawad et al . (2022) [16] observed that MRI offered more
Objective
evaluation of the junctional zone and better definition
of focal versus diffuse adenomyosis. They recommended MRI
when TVUS was inconclusive, when coexisting pathology was
suspected or when detailed mapping was needed before fertility -
preserving treatment. Harmsen et al. (2023) [17] emphasised that
the junctional zone is defined differently on TVUS, MRI and
histology. This lack of direct anatomical equivalence may
contribute to disagreement among modalities, even when
examinations are performed correctly.
Dason et al . (2023) [18] recommended TVUS as the initial
investigation for suspected adenomyosis and MRI as a second -
line modality when additional characterisation was required. The
present results support this sequential approach but also indicate
that MRI may be preferable when avoiding a missed diagnosis is
clinically important. The MRI positive likelihood ratio of 16.31
provided strong evidence for confirming adenomyosis, whereas
its negative likelihood ratio of 0.06 indicated that a negative
MRI result made the disease highly unlikely.
Mishra et al . (2023) [19] reported that MRI demonstrated high
concordance with histopathology, particularly when junctional -
zone measurements were combined with myometrial signal
abnormalities. Similarly, the present study recorded almost -
perfect MRI -histopathology agreement (κ=0.88). Nevertheless,
MRI should not be considered infallible because mild superficial
adenomyosis and small isolated foci may remain below spatial
resolution, while uterine contractions can sometimes simulate
focal junctional-zone thickening.
Agreement with histopathology and paired diagnostic
comparison
TVUS agreed with histopathology in 84.2% of cases and
demonstrated substantial agreement (κ=0.68). MRI agreed with
histopathology in 94.2% and demonstrated almost -perfect
agreement (κ=0.88). Agreement between TVUS and MRI was
86.7% (κ=0.73). Thus, the modalities generally classified the
same participants similarly, but MRI showed closer concordance
with the histopathological diagnosis.
In the paired analysis, MRI correctly classi fied 14 cases that
were incorrectly classified by TVUS, whereas TVUS correctly
classified only 2 cases missed by MRI. MRI detected an
additional 10.3% of histopathology -positive cases and
significantly reduced the false-negative rate from 16.2% to 5.9%
(p=0.039). Its improvement in specificity and false -positive rate
did not reach statistical significance, possibly because only 52
participants were histopathologically negative.
Kho et al. (2021) [20] described adenomyosis as a heterogeneous
condition for which histopathological confirmation itself may be
influenced by the number and location of tissue sections
examined. Consequently, imaging -histopathology disagreement
may partly reflect differences in sampling rather than imaging
failure alone. Kobayashi et al. (2021) [21] further demonstrated
that early and advanced pathological subtypes possess different
morphological appearances, potentially affecting detection by
both TVUS and MRI.
Munro (2019) [22] noted that imaging -based diagnosis has
become increasingly important because hysterectomy specimens
are unavailable in women undergoing conservative management.
The substantial agreement observed for TVUS supports its role
in routine screening and initial evaluation. Nevertheless, the
almost-perfect MRI agreem ent and significantly lower false -
negative rate suggest that MRI offers added value in women
with persistent symptoms, inconclusive TVUS, coexisting
leiomyomas, suspected focal disease or planned conservative
surgery.
Conclusion
Both transvaginal ultrason ography and magnetic resonance
imaging were effective non -invasive modalities for evaluating
adenomyosis. TVUS demonstrated good sensitivity, specificity
and substantial agreement with histopathology, supporting its
use as an accessible and cost -effective first-line investigation.
However, MRI showed significantly higher sensitivity and
overall diagnostic accuracy, stronger agreement with
histopathology and a lower false -negative rate. MRI may
therefore be particularly valuable when TVUS findings are
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 359 ~
inconclusive, coexisting uterine abnormalities are present or
accurate disease mapping is required. A sequential approach
involving initial TVUS followed by MRI in selected patients
may provide reliable and efficient diagnosis of adenomyosis.
Limitations
1. The study was conducted at a single tertiary -care hospital,
which may limit the generalisability of the findings to other
healthcare settings.
2. The sample size of 120 participants was relatively modest,
particularly for subgroup analyses based on focal or diffus e
adenomyosis.
3. Only women who underwent hysterectomy and had an
available histopathological diagnosis were included. This
may have introduced selection or verification bias and
increased the proportion of women with severe symptoms
or advanced disease.
4. The diagnostic performance observed in this surgical
population may not be directly applicable to younger
women, asymptomatic women or those receiving
conservative treatment.
5. TVUS is operator -dependent; therefore, its diagnostic
performance may have been i nfluenced by the experience
and expertise of the examining sonographer.
6. MRI interpretation may have been affected by variations in
menstrual-cycle phase, junctional -zone appearance,
transient uterine contractions and coexisting leiomyomas.
7. Complete blind ing between clinical information and
imaging findings might not have been possible in every
participant, creating a possibility of observer bias.
8. Interobserver and intraobserver variability in TVUS, MRI
and histopathological interpretation was not evaluated.
9. The diagnostic performances of individual MUSA
sonographic features and separate MRI criteria were not
analysed.
10. The study did not assess the cost -effectiveness,
accessibility, examination time or patient acceptability of
TVUS and MRI.
Conflict of Interest
Not available
Financial Support
Not available
References
1. Van den Bosch T, Dueholm M, Leone FPG, Valentin L,
Rasmussen CK, Votino A, et al . Terms, definitions and
measurements to describe sonographic features of
myometrium and uterine masses: a consensus opinion from
the Morphological Uterus Sonographic Assessment group.
Ultrasound Obstet Gynecol. 2015;46(3):284-298.
2. Pinzauti S, Lazzeri L, Tosti C, Centini G, Orlandini C, Luisi
S, et al . Transvaginal sonographic features of diffuse
adenomyosis in 18 -30-year-old nulligravid women without
endometriosis: association with symptoms. Ultrasound
Obstet Gynecol. 2015;46(6):730-736.
3. Bazot M, Daraï E. Role of transvaginal sonography and
magnetic resonance imaging in the diagnosis of uterine
adenomyosis. Fertil Steril. 2018;109(3):389-397.
4. Andres MP, Borrelli GM, Ribeiro J, Baracat EC, Abrão MS,
Kho RM. Transvaginal ultrasound for the diagnosis of
adenomyosis: systematic review and meta-analysis. J Minim
Invasive Gynecol. 2018;25(2):257-264.
5. Liu L, Li W, Leonardi M, Condous G, Da Silva Costa F,
Mol BW, et al . Diagnostic accuracy of transvaginal
ultrasound and magnetic resonance imaging for
adenomyosis: systematic review and meta-analysis. J
Ultrasound Med. 2019;38(12):3311-3324.
6. Tellum T, Nygaard S, Lieng M. Noninvasive diagnosis of
adenomyosis: a structured review and meta -analysis of
diagnostic accuracy in imaging. J Minim Invasive Gynecol.
2020;27(2):408-418.e3.
7. Chapron C, Vannuccini S, Santulli P, Abrão MS, Carmona
F, Fraser IS, et al. Diagnosing adenomyosis: an integrated
clinical and imaging approach. Hum Reprod Update.
2020;26(3):392-411.
8. Harmsen MJ, Van den Bosch T, de Leeuw RA, Dueholm
M, Exacoustos C, Valentin L, et al. Consensus on revised
definitions of Morphological Uterus Sonographic
Assessment features of adenomyosis: results of a modified
Delphi procedure. Ultrasound Obstet Gynecol.
2022;60(1):118-131.
9. Vannuccini S, Tosti C, Carmona F, Huang SJ, Chapron C,
Guo SW, et al. Pathogenesis of adenomyosis: an update on
molecular mechanisms. Reprod Biomed Online.
2017;35(5):592-601.
10. Abbott JA. Adenomyosis and abnormal uterine bleeding:
review of the evidence. Best Pract Res Clin Obstet
Gynaecol. 2017;40:68-81.
11. Exacoustos C, Morosetti G, Conway F, Camilli S, Martire
FG, Lazzeri L, et al . New sonographic classification of
adenomyosis: do type and degree of adenomyosis correlate
to severity of symptoms? J Minim Invasive Gynecol.
2020;27(6):1308-1315.
12. Tellum T, Matic GV, Dorm agen JB, Nygaard S, Viktil E,
Qvigstad E, et al . Diagnosing adenomyosis with MRI: a
prospective study revisiting the junctional -zone-thickness
cutoff of 12 mm as a diagnostic marker. Eur Radiol.
2019;29(12):6971-6981.
13. Rasmussen CK, Hansen ES, Dueholm M. In ter-rater
agreement in the diagnosis of adenomyosis by two - and
three-dimensional transvaginal ultrasonography. Ultrasound
Obstet Gynecol. 2019;54(4):512-518.
14. Van den Bosch T, de Bruijn AM, de Leeuw RA, Dueholm
M, Exacoustos C, Valentin L, et al . Sonograph ic
classification and reporting system for diagnosing
adenomyosis. Ultrasound Obstet Gynecol. 2019;53(5):576 -
582.
15. Agostinho L, Cruz R, Osório F, Alves J, Setúbal A, Guerra
A. MRI for adenomyosis: a pictorial review. Insights
Imaging. 2017;8(6):549-556.
16. Moawad G, Fruscalzo A, Youssef Y, Kheil M, Tawil T,
Nehme J, et al . Adenomyosis: an updated review on
diagnosis and classification. J Clin Med. 2022;11(16):4828.
17. Harmsen MJ, Trommelen LM, de Leeuw RA, Tellum T,
Juffermans LJM, Van den Bosch T, et al. Uterine junctional
zone and adenomyosis: comparison of MRI, transvaginal
ultrasound and histology. Ultrasound Obstet Gynecol.
2023;62(1):42-60.
18. Dason ES, Maxim M, Sanders AP, Papillon -Smith J, Thiel
JA. Guideline No. 437: diagnosis and management of
adenomyosis. J Obstet Gynaecol Can. 2023;45(6):417 -
429.e1.
19. Mishra P, Dash S, Jena SK, Behera S. Comparison of
magnetic resonance imaging and transvaginal
ultrasonography in the diagnosis of adenomyosis with
histopathological correlation. J Obstet Gynaecol India.
International Journal of Clinical Obstetrics and Gynaecology https://www.gynaecologyjournal.com
~ 360 ~
2023;73(4):342-349.
20. Kho KA, Chen JS, Halvorson LM. Diagnosis, evaluation,
and treatment of adenomyosis. JAMA. 2021;326(2):177 -
178.
21. Kobayashi H, Matsubara S, Imanaka S. Clinicopathological
features of different subtypes in adenomyosis: focus on
early lesions. PLoS One. 2021;16(7).
22. Munro MG. Classification and reporting systems for
adenomyosis. J Minim Invasive Gynecol. 2019;26(2):197 -
199.
How to Cite This Article
Reddy NA, Alluri DR, Reddy GH. Comparison of transvaginal
ultrasonography and magnetic resonance ima ging in the evaluation of
adenomyosis: A comparative study . International Journal of Clinical
Obstetrics and Gynaecology. 2026;10(5):352-360.
Creative Commons (CC) License
This is an open access journal, and articles are distributed under the terms
of the C reative Commons Attribution -Non-Commercial-Share Alike 4.0
International (CC BY -NC-SA 4.0) License, which allows others to remix,
tweak, and build upon the work non -commercially, as long as appropriate
credit is given and the new creations are licensed under the identical terms.
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.