基于T2WI肿瘤最大层面直方图分析鉴别诊断子宫内膜间质肉瘤与变性子宫肌瘤

Histogram analysis of the maximum level of tumor based on T2WI for differentiating endometrial stromal sarcoma from uterine degenerative leiomyoma

目的探讨基于T2WI肿瘤最大层面直方图分析鉴别诊断子宫内膜间质肉瘤(ESS)与变性子宫肌瘤的价值。方法回顾性分析20例ESS(ESS组)及24例变性子宫肌瘤(变性子宫肌瘤组)。采用MaZda软件于轴位脂肪抑制T2WI肿瘤最大层面勾画ROI,行灰度直方图分析,比较组间直方图参数差异;绘制受试者工作特征(ROC)曲线,评价直方图参数鉴别ESS与变性子宫肌瘤的效能。结果ESS组与变性子宫肌瘤组间轴位脂肪抑制T2WI肿瘤最大层面灰度直方图参数平均值、偏度及第1、10、50、90、99百分位数差异均有统计学意义(P均<0.05),方差及峰度差异无统计学意义(P均>0.05)。ROC曲线结果显示,上述组间差异有统计学意义的7个直方图参数中,以第50百分位数的曲线下面积(AUC)最大,为0.742(P=0.01),其鉴别诊断ESS与变性子宫肌瘤的敏感度为70.0%,特异度为83.3%。结论基于T2WI肿瘤最大层面直方图分析可鉴别ESS与变性子宫肌瘤。

Objective To explore value of histogram analysis of the maximum level of tumor on T2WI for differential diagnosis of endometrial stromal sarcoma(ESS)and degenerative uterine fibroids.Methods MRI data of 20 patients with ESS(ESS group)and 24 patients with degenerative uterine fibroids(degeneration fibroids group)were retrospectively analyzed.MaZda software was used to select ROI in the maximum level of tumor on axial lipid suppression T2WI,gray-scale histogram analysis was carried out to obtain characteristic parameters,and the parameters were then compared between groups.Receiver operating characteristic(ROC)curve was drawn to evaluate the efficacy of histogram parameters for distinguishing ESS from degeneration uterine fibroids.Results The mean,skewness,first percentile(per 1%),per 10%,per 50%,per 90%and per 99%were significant different between groups(all P<0.05),while no difference of variance nor kurtosis was found between groups(both P>0.05).ROC curve results showed that among the above histogram parameters,the maximum area under the curve(AUC)of the per 50%was 0.742(P=0.01),the sensitivity of differential diagnosis of ESS and degenerative uterine fibroids was 70.0%,and the specificity of 83.3%.Conclusion Histogram analysis of the maximum level of tumor based on T2WI could be used to distinguish ESS from degenerative uterine fibroids.