K Sasaguri1, K Yamaguchi2, T Nakazono2, M Mizuguchi2, S Aishima3, M Yokoyama4, H Irie2. 1. Department of Radiology, Faculty of Medicine, Saga University, Saga, 849-8501, Japan. Electronic address: k.sasaguri@gmail.com. 2. Department of Radiology, Faculty of Medicine, Saga University, Saga, 849-8501, Japan. 3. Department of Pathology and Microbiology, Faculty of Medicine, Saga University, Saga, 849-8501, Japan. 4. Department of Obstetrics & Gynecology, Faculty of Medicine, Saga University, Saga, 849-8501, Japan.
Abstract
AIM: To evaluate the accuracy of the ADNEX MR SCORING system for characterising adnexal masses. MATERIALS AND METHODS: An institutional review board approved this retrospective study. The study population comprised 663 women who underwent magnetic resonance imaging (MRI) from January 2007 to December 2014 to characterise 778 adnexal masses that were indeterminate under ultrasonography (590 benign and 188 malignant). Two radiologists independently reviewed the MRI images. The masses were scored from 1 to 5 according to the ADNEX MR SCORING system. The diagnostic performance of the system was evaluated by receiver operating characteristic (ROC) analysis. Masses scored 4 or greater were considered malignant (including tumours of borderline malignancy or low malignant potential). RESULTS: The malignancy rates of masses with scores of 2, 3, 4 and 5 were 1.9% (8/419), 12.8% (19/149), 62.6% (57/91) and 87.4% (104/119) for reader 1 and 2.1% (9/424), 13.6% (20/147), 67.6% (71/105) and 86.3% (88/102) for reader 2, respectively. The areas under the ROC curves for the differentiation of benign and malignant masses were 0.929 and 0.923, respectively; the sensitivity, specificity and accuracy of diagnosis were 85.6% (161/188), 91.7% (541/590), and 90.2% (702/778) for reader 1 and 84.6% (159/188), 91.9% (542/590), and 90.1% (701/778) for reader 2, respectively. Tumours of borderline malignancy or low malignant potential had a higher rate of misclassification (46.1%) than other malignant tumours (6-7.4%). CONCLUSION: The ADNEX MR SCORING system was highly accurate in differentiating benign and malignant adnexal masses, although it may be less accurate for tumours of borderline malignancy or low malignant potential.
AIM: To evaluate the accuracy of the ADNEX MR SCORING system for characterising adnexal masses. MATERIALS AND METHODS: An institutional review board approved this retrospective study. The study population comprised 663 women who underwent magnetic resonance imaging (MRI) from January 2007 to December 2014 to characterise 778 adnexal masses that were indeterminate under ultrasonography (590 benign and 188 malignant). Two radiologists independently reviewed the MRI images. The masses were scored from 1 to 5 according to the ADNEX MR SCORING system. The diagnostic performance of the system was evaluated by receiver operating characteristic (ROC) analysis. Masses scored 4 or greater were considered malignant (including tumours of borderline malignancy or low malignant potential). RESULTS: The malignancy rates of masses with scores of 2, 3, 4 and 5 were 1.9% (8/419), 12.8% (19/149), 62.6% (57/91) and 87.4% (104/119) for reader 1 and 2.1% (9/424), 13.6% (20/147), 67.6% (71/105) and 86.3% (88/102) for reader 2, respectively. The areas under the ROC curves for the differentiation of benign and malignant masses were 0.929 and 0.923, respectively; the sensitivity, specificity and accuracy of diagnosis were 85.6% (161/188), 91.7% (541/590), and 90.2% (702/778) for reader 1 and 84.6% (159/188), 91.9% (542/590), and 90.1% (701/778) for reader 2, respectively. Tumours of borderline malignancy or low malignant potential had a higher rate of misclassification (46.1%) than other malignant tumours (6-7.4%). CONCLUSION: The ADNEX MR SCORING system was highly accurate in differentiating benign and malignant adnexal masses, although it may be less accurate for tumours of borderline malignancy or low malignant potential.
Authors: Elizabeth A Sadowski; Isabelle Thomassin-Naggara; Andrea Rockall; Katherine E Maturen; Rosemarie Forstner; Priyanka Jha; Stephanie Nougaret; Evan S Siegelman; Caroline Reinhold Journal: Radiology Date: 2022-01-18 Impact factor: 11.105
Authors: Isabelle Thomassin-Naggara; Edouard Poncelet; Aurelie Jalaguier-Coudray; Adalgisa Guerra; Laure S Fournier; Sanja Stojanovic; Ingrid Millet; Nishat Bharwani; Valerie Juhan; Teresa M Cunha; Gabriele Masselli; Corinne Balleyguier; Caroline Malhaire; Nicolas F Perrot; Elizabeth A Sadowski; Marc Bazot; Patrice Taourel; Raphaël Porcher; Emile Darai; Caroline Reinhold; Andrea G Rockall Journal: JAMA Netw Open Date: 2020-01-03