The volumetric ADC histogram analysis in differentiating stage IA endometrial carcinoma from endometrial polyp.

Abstract

This study aimed to explore the value of ADC histogram based on whole lesion volume in distinguishing stage IA endometrial carcinoma from the endometrial polyp. MR images of 108 patients with endometrial lesions confirmed by pathology were retrospectively analyzed, including 65 cases of stage IA endometrial carcinoma and 43 cases of endometrial polyp. The volumetric ADC histogram metrics and general imaging features were evaluated and measured simultaneously. All the features were compared between the two groups. The receiver operating characteristic curve (ROC) was utilized to evaluate the diagnostic performance. The mean, max, min, and percentiles (10th, 25th, 50th, 75th, 95th) ADC values of endometrial carcinoma were significantly lower than that of polyp (all P<0.05). The skewness and kurtosis of ADC values in the endometrial carcinoma group were significantly higher than those in the endometrial polyp group, and the variance of ADC values in the endometrial carcinoma group was lower than those in the endometrial polyp group (all P<0.05). Endometrial carcinoma demonstrated more obvious myometrial invasion combined with intra-lesion hemorrhage than polyp (all P<0.05). The 25th percentile of ADC values achieved the largest AUC (0.861) among all the ADC histogram metrics and general imaging features, and the sensitivity and specificity were 83.08% and 76.74%, with the cut-off value of 1.01 × 10-3mm2/s. The volumetric ADC histogram analysis was an effective method in differentiating endometrial carcinoma from an endometrial polyp. The 25th percentile of ADC values has satisfactory performance for detecting malignancy in the endometrium. The ADC histogram metric based on whole lesion is a promising imaging-maker in differentiating endometrial benign and malignant lesions.