【摘要】 目的 探讨基于体素内不相干运动扩散加权成像(IVIM-DWI)的多指数模型对卵巢肿瘤的良恶性鉴别及与卵巢恶性肿瘤风险算法(ROMA)指数相关性分析。方法 共收集卵巢良性肿瘤39例与恶性肿瘤30例，临床和影像学资料均完整。基于IVIM-DWI多指数模型参数比较卵巢良恶性肿瘤的组间差异及各参数效能分析；再对上皮性卵巢癌(22例)IVIM-DWI各参数值与ROMA指数进行Spearman相关性分析。结果 卵巢良恶性肿瘤的IVIM-DWI多指数模型各定量参数组间比较差异显著，结果均有统计学意义(P<0.05)；其中Dslow、ADCstand及f对卵巢良恶性肿瘤的鉴别具有较高的诊断效能敏感性分别为86.67%,90%,72.33%，特异性分别为82.05%,87.18%,92.31%；上皮性卵巢癌ADCstand及Dslow值与ROMA指数呈负相关。结论 IVIM-DWI多指数模型联合ROMA指数有助于卵巢肿瘤的良恶性鉴别，可作为评估上皮性卵巢癌的有效生物学指标。更多还原

【Abstract】 Objective To investigate the multi-exponential model based on intra-voxel incoherent motion diffusionweighted imaging for the differentiation of benign and malignant ovarian tumors and the correlation analysis with the risk algorithm index of ovarian malignancy. Methods A total of 39 cases of benign ovarian tumors and 30 cases of malignant tumors were collected, and the clinical and imaging data were complete. Based on the IVIM-DWI multi-exponential model parameters, the differences between groups of ovarian benign and malignant tumors and the performance analysis of each parameter were compared. Then, the Spearman correlation analysis was performed on the IVIM-DWI parameters of epithelial ovarian cancer and the ROMA index. Results The quantitative parameters of IVIM-DWI multi-exponential model of ovarian benign and malignant tumors were significantly different between groups, and the results were all statistical y significant(P<0.05). The efficacy sensitivity was 86.67%, 90%, 72.33%, and the specificity was 82.05%, 87.18%, 92.31%, respectively. The ADCstand and Dslow values of epithelial ovarian cancer were negatively correlated with the ROMA index. Conclusion IVIM-DWI multi-exponential model combined with ROMA index can help differentiate benign and malignant ovarian tumors, and can be used as an effective biological indicator for evaluating epithelial ovarian cancer.更多还原