Diagnostic Accuracy of Transvaginal Ultrasound in Adenomyosis Taking MRI as a Gold Standard

Objective: To evaluate the accuracy of transvaginal ultrasound in the diagnosis of adenomyosis using MRI as the gold standard, and to characterise the most commonly seen and accurate ultrasonographic features and their combination. Study Design: Cross-sectional, descriptive study. Place and Duration of the Study: Department of Radiology, The Aga Khan University Hospital, Karachi, from January 2018 to July 2021. Methodology: Transvaginal ultrasound examination was performed on patients (n = 208) who presented with symptoms related to menstrual cycles and pelvic abnormalities. Additionally, patients who sought infertility evaluation were also included in the study. The findings from the ultrasound examinations were assessed and tabulated alongside the results of the MRI scans. All examinations were conducted by senior radiologists / sonographers. The sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively) of ultrasound features were calculated individually and in combination, taking MRI as the gold standard. To enhance the accuracy of ultrasound findings, various variables were combined, and their sensitivities and specificities were calculated. Results: Overall, transvaginal ultrasound had a high specificity of 96.15% (95% CI: 85.67 - 99.33), a relatively low sensitivity of 74.36% (95% CI: 66.63 - 80.85), PPV of 98.31 (95% CI: 93.40 - 99.70) and NPV of 55.56 (95% CI: 44.73 - 65.90). The most sensitive dual variable used was a bulky uterus combined with altered myometrial echotexture, with a sensitivity of 72.97% (95% CI: 64.95 - 79.78) and specificity of 95.83% (95% CI: 84.57-99.27). The best combined triple variable was a bulky uterus with altered echotexture and streaky myometrium, with a sensitivity of 71.85% (95% CI: 63.35 - 79.10) and a specificity of 95.46% (95% CI: 83.30 - 99.21). Conclusion: Transvaginal ultrasound features can identify adenomyosis characteristics in most of the patients. This could reduce the number of pelvic MRIs performed for the detection of adenomyosis.