Diagnostic Accuracy of Contrast-Enhanced MRI in Detection of Ovarian Malignant Masses

Abstract

Characterizing indeterminate adnexal masses remains a clinical challenge. Contrast-enhanced MRI (CE-MRI) may enhance the discrimination between benign and malignant ovarian lesions, thereby refining surgical triage. Objective: To determine the diagnostic accuracy of CE-MRI for detecting ovarian malignancy using histopathology as the reference standard. Methods: We conducted an observational study over six months, commencing from June 2024 to December 2024, at a tertiary referral center (Department of Radiology, JPMC, Karachi). Consecutive women (≥18 years) with suspected ovarian masses underwent standardized pelvic CE-MRI (multiparametric protocol with dynamic contrast enhancement). Imaging features (size, shape, septations/solid components, enhancement pattern, necrosis) and an overall CE-MRI impression (benign vs malignant) were recorded. All patients proceeded to definitive surgery; histopathology (malignant vs benign, subtype, grade) served as the Gold standard. Diagnostic performance metrics (sensitivity, specificity, PPV, NPV, accuracy) were calculated from 2×2 tables; ROC analysis quantified discrimination. Results: A total of 152 women were included (mean age 47.05 ± 11.28 years; symptom duration 8.69 ± 4.11 weeks). Mean lesion size was 8.54 ± 3.31 cm on baseline imaging and 7.93 ± 3.12 cm on CE-MRI. CE-MRI impressions were malignant in 75 (49.3%) cases; histopathology confirmed malignancy in 73 (48.0%). The 2×2 table yielded 66 true positives, nine false positives, 70 true negatives, and seven false negatives. CE-MRI demonstrated sensitivity 90.4%, specificity 88.6%, PPV 88.0%, NPV 90.9%, and overall accuracy 89.5%. ROC analysis showed strong discrimination (AUC 0.895, p < 0.0001). Conclusion: In a real-world referral cohort, CE-MRI demonstrated high diagnostic accuracy in differentiating malignant from benign ovarian masses, with excellent sensitivity and specificity, and an AUC approaching 0.90. CE-MRI thus provides robust second-line problem-solving after initial imaging and supports informed surgical planning.