Diagnostic Accuracy of Transvaginal Ultrasound in Detecting Uterine Adenomyosis Taking Magnetic Resonance Imaging as Gold Standard

To determine the diagnostic accuracy of transvaginal ultrasound (TVUS) in diagnosing uterine adenomyosis, using magnetic resonance imaging (MRI) as a gold standard. Study Design: A cross-sectional study. Place and Duration of the Study: Department of Radiology, Aga Khan University Hospital, Karachi, Pakistan, from January to June 2025. Methodology: A total of 231 women, between 20 and 50 years of age, presenting with heavy menstrual bleeding, dysmenorrhoea, irregular cycles, or chronic pelvic pain, were enrolled through consecutive sampling. All patients underwent TVUS followed by MRI within six months. Patients with malignancy, prior endometrial intervention, hysterectomy, or other confounding gynaecological conditions were excluded. The diagnostic accuracy indices of TVUS were calculated in comparison to MRI findings. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated.

Majority of the participants were between 20 and 50 years of age (87%), with 81.3% married and 39.1% multiparous. Heavy menstrual bleeding (51.7%) was the most common symptom, followed by dysmenorrhoea (20.4%) and chronic pelvic pain (17.4%). TVUS was positive in 52% of cases, and MRI confirmed adenomyosis in 87.4%. Compared with MRI, TVUS showed a sensitivity of 56.9%, specificity of 82.8%, PPV of 95.8%, NPV of 21.6%, and overall diagnostic accuracy of 60.2%.

TVUS demonstrated high specificity and PPV but limited sensitivity in diagnosing adenomyosis compared to MRI. It remains a valuable first-line diagnostic tool in low-resource settings, while MRI continues to serve as the reference standard for confirmatory evaluation. Key Words: Adenomyosis, Transvaginal ultrasound, Magnetic resonance imaging, Diagnostic accuracy, Heavy menstrual bleeding, Dysmenorrhoea.

Adenomyosis is a benign gynaecological disorder characterised by the presence of endometrial glands and stroma within the myometrium.1 It typically manifests with heavy menstrual bleeding and dysmenorrhoea, while some patients may also present with chronic pelvic pain.2 The true prevalence of adenomyosis remains difficult to establish, with reports ranging between 5% and 70%.3 The condition is most commonly encountered in multiparous women during the late reproductive years.1 Histopathologically, adenomyosis can be categorised into diffuse and focal types.4 In diffuse adenomyosis, the uterus appears enlarged and boggy with a thickened myometrial wall containing small haemorrhagic foci. Conversely, focal adenomyosis, also termed adenomyoma, may mimic a fibroid but lacks the typical pseudocapsule.5,6 The primary imaging modality for evaluating adenomyosis is transvaginal ultrasound (TVUS).3 However, magnetic resonance imaging (MRI) remains the reference standard, particularly to differentiate between diffuse and focal adenomyosis.5 It also helps to differentiate an adenomyoma from a leiomyoma. A previous systematic review in 2007 reported a sensitivity of 74%, specificity of 87%, positive predictive value (PPV) of 68%, and negative predictive value (NPV) of 89% for TVUS in diagnosing adenomyosis.7 More recently, a meta-analysis by Liu et al. demonstrated that TVUS had higher sensitivity (81%) compared with MRI (71%), while MRI was superior in specificity (91% vs. 71% for TVUS).3 In low-resource settings such as Pakistan, ultrasound offers practical advantages: it is more affordable, readily available, and less time-consuming compared to MRI. This underscores the importance of assessing the diagnostic accuracy of TVUS in local clinical practice. Although histopathology is the definitive diagnostic method, its use is often limited since many women with adenomyosis are of reproductive age and may not undergo hysterectomy. Additionally, given the benign nature of the disease, some patients prefer conservative management. For these reasons, MRI serves as a non-invasive confirmatory tool. Champaneria et al. reported MRI to have a sensitivity of 77%, specificity of 89%, positive likelihood ratio of 6.5, and negative likelihood ratio of 0.2 for diagnosing adenomyosis.8 Although previous studies have frequently compared TVUS against histopathology, the clinical reality is that histological confirmation is often impractical. This creates a strong rationale for evaluating TVUS against MRI, which, despite being non-invasive, can reliably serve as the gold standard in patients who cannot undergo surgical confirmation.9-11 The objective of this study was to determine the diagnostic accuracy of TVUS in detecting uterine adenomyosis, using MRI as the gold standard. METHODOLOGY This is a cross-sectional study carried out at the Department of Radiology, Aga Khan University Hospital, Karachi, Pakistan, from January to June 2025. Ethical approval was obtained from the hospital’s Institutional Review Board prior to starting the study. A sample size of 231 patients was calculated based on sensitivity and specificity values reported in previous studies,2 using a 95% confidence interval and a 9% margin of error. A non- probability consecutive sampling technique was used for patient selection. Females between 20 and 50 years of age presenting with dysmenorrhoea, heavy menstrual bleeding, irregular cycles, and/or chronic pelvic pain were included. However, few women over the age of 50 years were also included. Only those patients who had undergone TVUS at the study centre and subsequently underwent MRI within six months were considered eligible. Patients who gave a history of malignancy, transcervical resection of the endometrium, indications for hysterectomy, or those showing features of adenomyosis on TVUS but not undergoing MRI were excluded. Similarly, patients who had received any form of treatment for adenomyosis or who suffered from other gynaecolo- gical illnesses were also excluded to minimise confounding factors. Prior surgical history was confirmed by asking the patients and reviewing their available medical records. All patients fulfilling the inclusion criteria underwent a detailed clinical assessment followed by TVUS, which was performed by a radiologist or a sonologist. Patients who were advised to get an MRI of the pelvis for further evaluation and subsequently underwent the scan were included in the final analysis. Both TVUS and MRI findings were retrieved through the hospital PACS system, with strict confidentiality maintained regarding patient identity and the reporting physician. Reports were anonymised, and data were recorded on a pre-designed proforma, which included demographic details such as age, marital status, and parity, as well as presenting symptoms, history of pelvic disease, and imaging findings. Data were analysed using SPSS version 26. Categorical variables, including clinical features, parity, prior pelvic surgery, and imaging outcomes, were expressed as frequencies and percentages. A 2 × 2 contingency table was generated against MRI results by using true and false positives and true and false negatives. Sensitivity, specificity, PPV, and NPV were then calculated to determine the diagnostic accuracy of TVUS. A p-value of <0.05 was considered statistically significant.

Out of the 231 cases, 87.0% were between 20 and 50 years of age, while only 13.0% were older than 50 years. Most of the patients were married (81.3%), while 18.7% were single. 60.9% of the sample population had 0-2 children, while 39.3% had more than two children. Patients presenting with heavy menstrual bleeding were defined as those who had blood loss exceeding typical menstrual volume: menstrual periods lasting more than seven days; bleeding heavy enough to require changing a pad or tampon every hour for several consecutive hours; the need to use multiple pads simultaneously to prevent leakage; waking at night to change sanitary products; passing large blood clots multiple times a day; or being unable to carry out normal daily activities during menstruation. Patients with irregular cycles were those whose menstruation would start before 21 days or after 35 days since the start of the previous cycle. Those presenting with pelvic pain were patients who had pelvic pain lasting more than 6 months. More than half of patients presented with heavy menstrual bleeding, while dysmenorrhoea, irregular cycles, and chronic pelvic pain were less frequent. Heavy menstrual bleeding was reported by 51.5% of the patients, while 48.4% of the study population did not suffer from heavy menstrual bleeding. Similarly, 20.3% of the population had complaints of dysmenorrhea while 79.6% did not. Irregular cycles were seen in 16.8% of the patients, with 83.1% having no such complaints. Chronic pelvic pain was also not commonly seen, with only 17.4% of the population presenting with such complaints while 82.6% had no such complaints. Out of the 231 patients, 28 (12.2%) had a history of pelvic surgery, while the majority (87.8) did not. On diagnostic evaluation, TVUS showed positive findings in 52.0% of cases and negative results in 48.0%. Similarly, MRI revealed positive findings in 87.4% and negative findings in 12.6% of patients (Table I). Using MRI as the gold standard, the diagnostic accuracy of TVUS was evaluated in 231 cases. Of the 202 MRI-confirmed positives, TVUS correctly detected 115 (49.8%) but failed to identify 87 (37.7%). Table I: Diagnostic findings (n = 231).| Variables | Categories | n | Percentage (%) | | TVUS | Positive | 120 | 52.0 | | Negative | 111 | 48.0 | | | MRI | Positive | 202 | 87.4 | | Negative | 29 | 12.6 | Table II: Diagnostic accuracy of TVUS using MRI as gold standard (n = 231). | Variables | MRI positive | MRI negative | Total | | TVUS positive | 115 (49.8%) | 5 (2.2%) | 120 (52.0%) | | TVUS negative | 87 (37.7%) | 24 (10.4%) | 111 (48.0%) | | Total | 202 (87.5%) | 29 (12.5%) | 231 (100%) | Table III: Findings seen in patients with TVUS positive for adenomyosis (n = 120). | TVUS findings | n | Percentage (%) | | Heterogenous myometrium | 115 | 96 | | Thickened or indistinct junctional zone | 12 | 10 | | Sub-endometrial cysts | 12 | 10 | | Sub-endometrial linear echogenic striations | 7 | 5.8 | | Increased number of vessels seen penetrating the myometrium | 4 | 3.3 | | Sub-endometrial small hypoechoic nodules | 0 | 0 | Table IV: Findings seen in patients with MRI positive for adenomyosis (n = 202). | MRI | n | Percentage (%) | | Diffusely thickened junctional zone | 169 | 84 | | Thickened and heterogenous myometrium | 24 | 11.9 | | Focal widening of junctional zone | 19 | 9.4 | | T2 hyperintense signals representing cystic areas | 78 | 38.6 | | T1 hyperintense signals representing hemorrhage into ectopic endometrial tissues | 23 | 11.4 | | T1 post contrast enhancement of ectopic endometrial glands | 0 | 0 | Among the 29 MRI-negative cases, TVUS accurately classified 24 (10.4%) as negative, while 5 (2.2%) were misclassified as false positives. This corresponded to a sensitivity of 56.9%, specificity of 82.8%, PPV of 95.8%, and NPV of 21.6%. Overall, the diagnostic accuracy of TVUS relative to MRI was 60.2% (Table II). Out of the 120 patients with positive TVUS for adenomyosis, the most commonly seen finding was heterogeneity of the myometrium, which was seen in 96% of the patients (Figure 1). The rest of the findings were less commonly seen. Percentages for all the mentioned findings are given in Table III. Multiple findings could be seen in one patient. Out of the 202 patients with positive MRI for adenomyosis, the most commonly seen finding was diffusely thickened junctional zone in 84% of the patients, followed by T2 hyperintense signals representing cystic areas, which were seen in 38.6% of the patients (Figure 2). The rest of the findings were less commonly seen. Percentages for all the above-mentioned findings are given in Table IV. Similar to TVUS, multiple findings could be seen in one patient. Figure 1: TVUS was performed in a 45-year with complaints of heavy menstrual bleeding. The ultrasound reveals a heterogeneous appearance of the myometrium with multiple cystic spaces, as depicted by the white arrows. Figure 2: MRI of the pelvis of the same 45-year patient (Figure 1) was subsequently performed, which showed diffusely thickened myomet-rium with multiple cystic spaces appearing hyperintense on T2-weighted images. The most commonly seen finding on TVUS, which was hetero- geneity of the myometrium, corresponded to the most commonly seen finding in MRI, which was a diffusely thickened junction zone, in most cases. In some patients, heterogeneity of the myometrium corresponded to thickened and heterogeneous myometrium on MRI. For the five cases that were positive for adenomyosis on TVUS but negative on MRI, four patients were found to have heterogeneous myometrium on TVUS, and one was found to have a thickened or indistinct junctional zone. For these five cases, two patients had chronic pelvic pain, two had irregular cycles, and only one had heavy menstrual bleeding.

This study assessed the diagnostic performance of TVUS for adenomyosis, using MRI as the reference standard. The results indicate that TVUS is highly specific and reliable when positive, but its relatively low sensitivity and NPV limit its ability to rule in disease. Several recent studies have explored the diagnostic accuracy of ultrasound in comparison with MRI and histopathology. Hashad et al. found that three-dimensional ultrasonography performed with comparable accuracy to MRI in diagnosing adenomyosis and leiomyomas, emphasising its accessibility and cost-effectiveness in routine clinical practice.11 Similarly, Anwar et al. compared TVUS and MRI against histopathology, reporting higher diagnostic accuracy for MRI (78.4%) compared to TVUS (60.5%). Their findings align with the results of the present study in demonstrating the superior performance of MRI but also confirm the role of TVUS as a useful frontline tool.10 Contrary to the relatively modest sensitivity (56.9%) found in this study, Shehzadi et al. demonstrated much higher diagnostic performance of TVUS, with a sensitivity of 94.1% and specificity of 89.8% compared with MRI.12 This variation may be attributed to operator expertise, differences in imaging criteria, and study settings. Likewise, Shaikh et al. also reported higher specificity (96.1%) and PPV (98.3%) for TVUS, although their sensitivity (74.3%) was closer to the sensitivity in this study, again showing that false negatives remain a challenge.13 Studies comparing both MRI and ultrasound with histopathology also demonstrate heterogeneity. Anjum et al. found higher sensitivity for sonography (88.0%) than MRI (84.0%), but MRI outperformed in specificity (96.0% vs. 92.1%) and overall accuracy (93.0% vs. 91.0%).14 Similarly, Elsaid et al. compared MRI and ultrasound for multiple uterine lesions, reporting that MRI had superior sensitivity in detecting malignancy, while USG performed adequately for benign conditions such as adenomyosis.15 These findings support the complementary role of both modalities, with ultrasound offering a practical screening tool and MRI providing more definitive evaluation when required. Meta-analyses and narrative reviews also corroborate this variability. Gallo et al. emphasised that TVUS may achieve higher sensitivity, but MRI has greater specificity.16 Hansen et al., although focusing on deep infiltrating endometriosis, also highlighted the comparable accuracy of TVUS and MRI, noting that performance often depends on operator skill and lesion location.17 The variability across studies underscores the influence of study design, patient characteristics, and radiologist exper- tise.18-20 In this study, variability in operator skills, which includes years of experience with some operators in their early years of sonographic work, while others have vast experience, may have contributed to the relatively low sensitivity observed. From a clinical perspective, the high PPV (95.8%) in this study suggests that positive TVUS findings can reliably guide patient management in resource-limited settings, reducing the need for routine MRI. However, the low NPV highlights the risk of underdiagnosis if TVUS is solely relied upon, necessitating MRI in patients with persistent symptoms despite negative ultrasound findings. The most commonly seen finding on TVUS, which was heterogeneity of the myometrium, corresponded to the most commonly seen finding in MRI, which was a diffusely thickened junction zone, in most cases. This suggests that these two findings in their respective studies will be seen commonly and will be most frequently associated with patients with positive imaging for adenomyosis. As previously mentioned in the results section, for the five cases that were positive for adenomyosis on TVUS but negative on MRI, four patients were found to have heterogeneous myometrium on TVUS, and one was found to have a thickened or indistinct junctional zone. This may reflect the variability in operator skill level. There are some limitations to this study. First, certain data points—such as patient-reported symptoms and surgical history—were self-reported and may be subject to recall or reporting bias. Second, variability in operator skill levels may have influenced the imaging results, potentially affecting the consistency and reliability of the findings.

The findings of this study reaffirm that while MRI remains the gold standard imaging modality, TVUS continues to serve as a cost-effective and widely available diagnostic tool, especially valuable in low-resource settings such as Pakistan. However, efforts to improve diagnostic performance through operator training, adoption of 3-D ultrasound, and incorporation of standardised sonographic criteria may further enhance its utility in routine practice. ETHICAL APPROVAL: Ethical approval was obtained from the Institutional Review Board of the Aga Khan University Hospital, Karachi, Pakistan, prior to starting the study. PATIENTS’ CONSENT: Informed consent was obtained from all participants. COMPETING INTEREST: The authors declared no conflict of interest. AUTHORS’ CONTRIBUTION: MS, SAS: Conception and design of the work, acquisition, analysis, interpretation of data, drafting, and critical revision of the manuscript for important intellectual content. Both authors approved the final version of the manuscript to be published.

- Bazot M, Darai E. Role of transvaginal sonography and magnetic resonance imaging in the diagnosis of uterine adenomyosis. Fertil Steril 2018; 109(3):389-97. doi: 10. 1016/j.fertnstert.2018.01.024. - Rubab S, Maqsood A, Rafique MS, Kundi S, Yasi A, Tariq T. Comparing accuracies of transvaginal ultrasonography and magnetic resonance imaging for nonsurgical diagnosis of adenomyosis. J Soc Obstet Gynaecol Pak 2022; 12(3). - 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 and review of sonographic diagnostic criteria. J Ultrasound Med 2021; 40(11):2289-306. doi: 10.1002/jum.15635. - Dueholm M, Lundorf E, Hansen ES, Sorensen JS, Ledertoug S, Olesen F. Magnetic resonance imaging and transvaginal ultrasonography for the diagnosis of adenomyosis. Fertil Steril 2001; 76(3):588-94. doi: 10.1016/s0015- 0282(01) 01962-8. - Stewart EA. Uterine adenomyosis. Post TW, ed. UpToDate. Waltham, MA: UpToDate Inc. Available from: httpss://www. uptodate.com/contents/uterine-adenomyosis/print? source=related_link. - Kepkep K, Tuncay YA, Goynumer G, Tutal E. Transvaginal sonography in the diagnosis of adenomyosis: Which findings are most accurate? Ultrasound Obstet Gynecol 2007; 30(3):341-5. doi: 10.1002/uog.3985. - Dueholm M, Lundorf E. Transvaginal ultrasound or MRI for diagnosis of adenomyosis. Curr Opin Obstet Gynecol 2007; 19(6):505-12. doi: 10.1097/GCO.0b013e3282f1bf00. - Champaneria R, Abedin P, Daniels J, Balogun M, Khan KS. Ultrasound scan and magnetic resonance imaging for the diagnosis of adenomyosis: Systematic review comparing test accuracy. Acta Obstet Gynecol Scand 2010; 89(11): 1374-84. doi: 10.3109/00016349.2010.512061. - Gaillard F, Haouimi A, Bell D, et al. Adenomyosis. Reference article, Radiopaedia.org (Accessed on 09 May 2023). httpss://doi.org/10.53347/rID-10171 - Anwar J, Tariq M, Amin N, Gul B, Niaz A, Nawaz KH. Diagnostic accuracy of ultrasound and MRI for diagnosis of adenomyosis taking histopathology as gold standard. Pakistan Armed Forces Med J 2022; 72(SUPPL-2):S346-49. doi: 10.51253/pafmj.v72iSUPPL-2.8064. - Hashad AM, Hassan NE, Elbohoty AE, Bakr OB, Ibrahim IM. 3D ultrasonography compared with magnetic resonance imaging for the diagnosis of adenomyosis. Egypt J Hosp Med 2017; 69(8):3123-33. Available from: httpss://appli cations.emro.who.int/imemrf/Egypt_J_Hosp_Med/Egypt_J_Hosp_Med_2017_69_8_3123_3133.pdf. - Shehzadi Z, Jadoon F, Aquil H, Kayani A, Khan SA. Diagnostic accuracy of transvaginal ultrasonography in diagnosing adenomyosis, taking MRI findings as gold stan-dard. Pak J Radiol 2024; 34(1). - Shaikh A, Masroor I, Masood A, Saeed SA. Diagnostic accuracy of transvaginal ultrasound in adenomyosis taking MRI as a gold standard. J Coll Physicians Surg Pak 2023; 33(10):1118-23. doi: 10.29271/jcpsp.2023.10.1118. - Anjum A, Umbreen S, Nadeem MF, Noureen F, Asim F, Anwar S. Sonography and magnetic resonance imaging for the diagnosis of adenomyosis: Comparison with histopathology. J Islamabad Med Dent Coll 2024; 13(4). doi: 10.35787/jimdc.v13i4.1030. - Elsaid AY, Youssef NA, Khater HM. Comparative study between MRI and ultrasound in evaluation of uterine lesions. BMFJ 2025. doi 10.21608/bmfj.2025.344381.2286. - Gallo R, Kamberaj L, Baroni A, Nardone AD, Scambia G, Masciullo V. Advances in non-invasive diagnosis of uterine adenomyosis: A narrative review. Gynecol Pelvic Med 2025; 8. doi: 10.21037/gpm-24-52. - Hansen T, Hanchard T, Alphonse J. The accuracy of ultrasound compared to magnetic resonance imaging in the diagnosis of deep infiltrating endometriosis: A narrative review. Sonography 2023; 10(2):76-85. doi: 10.1002/sono. 1235. - Garcia-Solares J, Donnez J, Donnez O, Dolmans MM. Pathogenesis of uterine adenomyosis: invagination or metaplasia? Fertil Steril 2018; 109(3):371-9. doi: 10. 1016/j.fertnstert.2017.12.030. - Donnez J, Donnez O, Dolmans MM. Introduction: Uterine adenomyosis, another enigmatic disease of our time. Fertil Steril 2018; 109(3):369-70. doi: 10.1016/j.fertnstert. 2018.01.035. - Gordts S, Grimbizis I, Campo R. Symptoms and classifi-cation of uterine adenomyosis. Fertil Steril 2018; 109(3): 380-8.e1. doi: 10.1016/j.fertnstert.2018.01.006.