Synopsis

Keywords: Urogenital, MR Value, ovarian

The newly launched O-RADS scoring system was developed to address the qualitative problem of pelvic masses. However, the use of the O-RADS system in different regional populations still needs to be further verified. And the applicability of the 0-RADS scoring system for doctors of different seniority remains to be studied.

Purpose

To verify the accuracy of O-RADS-MRI scoring system in classifying benign or malignant adnexal lesions by mp-MRI, and inspect its applicability in doctors with different seniority. Try to simplify the O-RADS process, draw flowcharts and discuss the experience.

Method

The basic information of 343 patients with pelvic lesions who underwent surgery at our institution, was retrospectively analyzed. Exclusion criteria included previous removal of the uterus for other reasons, inability to establish time intensity curves for technical reasons, multiple phases enhanced imaging that did not contain adnexal lesions, adnexal masses that cannot be recognized by the naked eye, and tumors raising from non-adnexal origin. Eventually we targeted 305 patients, 350 lesions totally. We recorded their MR imaging and postoperative histopathology. A junior radiologist and senior radiologist scored each lesion according to O-RADS criteria. After fully comprehending the O-RADS White Paper, our team has created a simple flowchart of O-RADS for readers to reference. With histological diagnosis as the gold standard, making 4 as the cutoff value as O-RADS scoring system differentiating benign and malignant, to test the diagnostic accuracy of two radiologists. Due to lack of expert consensus on treatment, borderline ovarian tumors (BOTs) were classified as malignant in this study. Cohen's kappa coefficient was applied to calculate the inter-class agreement. ROC of each radiologist was depicted. A P < 0.05 was considered to be statistically significant.

Results

The whole group included 204 malignant and 146 benign lesions. Both raters had a good subjective experience with the simple flowchart and felt that the content of O-RADS could be better understood in a shorter period of time. Great agreement between readers was found (к =0.633, 95% confidence interval [CI]0.72~0.60). The results of specified threshold of malignancy of junior radiologist turns out a sensitivity of 0.858, a specificity of 0.973, and AUC was 0.932 (95% CI 0.90-0.96), and senior radiologist turns out a sensitivity of 0.922, a specificity of 0.897, [2] and AUC was 0.950 (95% CI 0.03-0.98).

Conclusion

Our study shows that O-RADS distinguish between benign and malignant well, and radiologists with different experience both have a good grasp of the O-RADS process. Although streaming the O-RADS scoring process has improved the scoring speed to a certain extent, but in precondition that mastering basic content of O-RADS is necessary.

Acknowledgements

Throughout this study I have received a great deal of support and assistance. I would first like to thank my supervisor, Chen Yan, whose expertise was invaluable in formulating the research questions and methodology. Your insightful feedback pushed me to sharpen my thinking and brought my work to a higher level. I would particularly like to acknowledge my teammate tutor, Xu Xiaojuan, for their wonderful collaboration and patient support Finally, I could not have completed this study without the support of my friends and colleagues, who provided stimulating discussions as well as happy distractions to rest my mind outside of my research.