Applying contrast-enhanced ultrasound model to distinguish atypical focal adenomyosis from uterine leiomyomas

Background: Different from conventional ultrasound, contrast-enhanced ultrasound (CEUS) is better in observing microperfusion. For atypical focal adenomyosis and uterine leiomyomas that are difficult to be distinguished by conventional ultrasound, this study aims to further improve the differential diagnosis performance by using CEUS model. Methods: After screening the cases with difficulties in identifying focal myometrium lesions through conventional ultrasound, the number of cases covered in the focal adenomyosis group and leiomyomas group were 60 and 30 in derivation cohort, 14 and 7 in validation cohort. The qualitative and quantitative characteristics of CEUS were analyzed according to the surgical pathology. The qualitative characteristics include: the enhancement level based on the myometrium, the contrast enhancement pattern, the enhanced time of the lesion based on the myometrium, post-contrast lesion border, the distribution of the contrast agent, and the wash-out time based on the myometrium. The quantitative characteristics include: arrive time (AT), time to peak (TTP), peak intensity (PI), Delta AT, Delta TTP, Delta PI, vertical bar Delta AT vertical bar, vertical bar Delta TTP vertical bar, vertical bar Delta PI vertical bar and lesion temporal variability. Multiple logistic regression analysis was used to screen the independent risk factors, and a risk prediction model for the differential diagnosis of the two diseases was established. The area under the receiver operating characteristic (ROC) curve (AUC) was used to assess the diagnostic performance of the model. The validation cohort was used to further evaluate the diagnostic performance of the model. Results: Through the multivariate analysis, it concluded that short-term vessels first enhanced enhancement mode, unclear boundary, lesion temporal variability under CEUS >9.5 s, uneven contrast agent distribution could be independent risk factors for the diagnosis of adenomyosis [AUC =0.908, 95% confidence interval (CI): 0.833-0.982]. We also determined the sensitivity (98.33%), specificity (70.00%), positive predictive value (PPV) (86.76%), negative predictive value (NPV) (95.45%), and accuracy (87.78%) of this model. Based on pathological diagnosis, the sensitivity and specificity of the model in the validation cohort were both 85.71%, with NPV of 75% and PPV of 92.3%. The area under the ROC curve was 0.898 (95% CI: 0.742-1.000). Conclusions: The establishment of CEUS model has certain clinical application value in differentiating atypical focal adenomyosis from leiomyomas.