Differentiating between benign and malignant ovarian lesions using magnetic resonance imaging assessment: A systematic review and diagnostic accuracy meta-analysis

Abstract

Author(s): Zeinab Safarpour Lima, Ayda Roostaee

Background and Aim: Clinical diagnosis of ovarian cancer involves a combination of symptoms, blood tumor marker tests, and MRI images. Accurate diagnosis is essential for developing effective treatment strategies. MRI is commonly used due to its convenience. This study aimed to assess the diagnostic value of preoperative MRI in distinguishing between benign and malignant ovarian lesions before surgery.

Methods: We performed a systematic search of literature in PubMed, Web of Science, and Scopus with relevant keywords. Studies that did not perform MRI or had insufficient data were excluded. Data extraction was performed based on a standardized sheet. Meta analysis was performed with STATA, R, and R-Studio.

Results: The initial search retrieved 14,967 articles from which 3,921 duplicates were removed. Finally, 15 studies were included based on our eligibility criteria. The pooled sensitivity of MRI in detection benign and malignant lesions was 89% (95% CI: 81%-94%, p-value<0.01). The pooled specificity MRI in detection benign and malignant lesions was 94% (95% CI: 90%-97%, p-value<0.01). There was considerable heterogeneity among the included studies. The I2 index indicates a generalized heterogeneity of 61% with heterogeneity of the sensitivity and specificity being 67% and 57%, respectively.

Conclusion: MRI shows high sensitivity, specificity, and diagnostic accuracy in distinguishing between benign and malignant ovarian tumors. Most studies reported sensitivity above 80% and specificity exceeding 90%. Further large-scale, multi-center prospective studies are needed to further evaluate the diagnostic efficacy of MRI in diagnosing ovarian neoplasms.

Share this article