The role of apparent diffusion coefficient histogram analysis in improving O-RADS magnetic resonance imaging classification of adnexal lesions: a retrospective diagnostic accuracy study

Background: Ovarian cancer is the fourth leading cause of death among neoplastic diseases in women. The Ovarian-Adnexal Reporting Data System (O-RADS) magnetic resonance imaging (MRI) scoring is a five-point risk stratification system that combines morphological features with diffusion and perfusion features, which allows categorization of adnexal masses. This study aimed to evaluate the value of apparent diffusion coefficient (ADC) histogram analysis and serum carbohydrate antigen 125 (CA-125) levels to improve the performance of the O-RADS MRI score in classification of adnexal lesions.

Methods: The retrospective study included 196 patients with 235 adnexal lesions. Exclusion criteria included lack of histopathological results, prior treatment for adnexal lesions, poor image quality, and absence of ADC maps or contrast-enhanced (CE) magnetic resonance (MR) images. The MRI protocol included T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), diffusion weighted imaging (DWI), and CE sequences, with ADC histogram parameters extracted using FireVoxel software. Histological results of surgically excised specimens served as the reference standard. Clinical information, ADC histogram parameters and O-RADS MRI scores were compared between benign and malignant lesions. Predictive factors for malignancy were identified through univariate and multivariate analyses, while diagnostic performance of the multivariate model, O-RADS score, and their combination was assessed using receiver operating characteristic (ROC) curve analysis.

Results: Significant differences in ADC histogram parameters were observed between benign and malignant lesions, except for the ADC maximum value (ADC_Max) (P=0.97). The O-RADS MRI score demonstrated strong diagnostic performance for malignancy with an area under the curve (AUC) of 0.915, a sensitivity of 84.25%, a specificity of 90.74%, and an accuracy of 87.23%. Incorporating ADC mean value (ADC_Mean), ADC Kurtosis (ADC_Kurtosis), and the serum CA-125 levels further improved differentiation, achieving a higher AUC (0.956), with a sensitivity of 92.91%, a specificity of 89.81%, and accuracy of 90.64%.

Conclusions: The ADC histogram analysis may improve the diagnostic performance of the O-RADS MRI score in preoperative classification of adnexal lesions.