MR扩散加权成像鉴别卵巢交界性上皮性肿瘤与卵巢癌的b值优化

Optimization of b Value of Diffusion-weighted Imaging for Discriminating between Borderline Epithelial Ovarian Tumors and Ovary Cancer

目的:探讨MR扩散加权成像(DWI)鉴别卵巢交界性上皮性肿瘤(BEOT)与卵巢癌的最佳b值。方法:DWI扫描采用平面回波成像(EPI)技术,扩散敏感因子(b)值分别取0s/mm2,150s/mm2,500s/mm2,800s/mm2,1000s/mm2。对52例经手术病理证实的BEOT与卵巢癌患者行MR DWI检查。其中,BEOT 27例32个肿瘤,卵巢癌25例33个肿瘤。测量DWI图像信噪比(SNR)及对比噪声比(CNR)、实性成分与囊性成分信号强度比值(R),ADC图上测量实性成分ADC值,两位观察者DWI图像上评价肿瘤实性成分信号。结果:随b值增大DWI图像SNR值及CNR值均下降。b值≥500s/mm2时,随b值增大实性成分与囊性成分DWI信号强度对比(R)增加。不同b值测得的肿瘤实性成分的ADC值在BEOT与卵巢癌间差异均有统计学意义。四组b值(150s/mm2、500s/mm2、800s/mm2及1000s/mm2)ROC曲线下面积(AUC)分别为0.934、0.976、0.979和0.979。b值=1000s/mm2时,两位观察者对肿瘤DWI信号评价仍具有一致性,绝大多数卵巢癌DWI呈强信号。结论:DWI鉴别卵巢交界性上皮性肿瘤与卵巢癌的最佳b值为1000s/mm2。

Purpose： To determine optimal b value of diffusion-weighted imaging （DWI） for discriminating borderline epithelial ovarian tumors and ovary cancer. Methods： Twenty seven patients with 32 borderline epithelial ovarian tumors （BEOT） and 25 patients with 33 ovary cancers underwent 1.5-T MRI with EPI- DWI （b=lS0s/mm^2, 500s/mm^2, 800s/mm^2, 1000s/romE）. Signal noise ratio （SNR） and contrast noise ratio （CNR） were calculated on DWI. Signal intensity of tumors was measured on DWI, and SIsolid/SIcystic （R） was calculated. ADC value was measured on ADC map. Signal intensity on DWI was assessed by two radiologists. Results： There was an inverse ratio between b value and SNR and CNR of DWI. However, a higher b value provided a higher contrast of DWI signal between solid and cystic components. ADC value of solid component was decreased along with the increase of b value. The b values of 800s/mm^2 and 1000s/mm^2 had the same AUC of 0.979 for discriminating BEOT and ovary cancer. There was a consistency between two radiologists in assesment of DWI signal with a b value of 1000s/mm^2. The majority of ovary cancer showed high signal intensity on DWI. Conclusion： A b value of 1000s/mm^2 is optimal value of DWI for discriminating borderline epithelial ovarian tumors and ovary cancer.