MRI多参数定量分析对鼻腔鼻窦内翻性乳头状瘤恶性转化的预测价值

Predictive value of MRI multi-parameter quantitative analysis for malignant transformation of sinonasal inverted papilloma

目的:探究MRI多参数定量分析对鼻腔鼻窦内翻性乳头状瘤(sinonasal inverted papilloma,SNIP)恶性转化的预测价值。方法:回顾性分析83例SNIP和21例SNIP转化鳞状细胞癌(SNIP-SCC)患者的临床及影像资料。采用单因素分析和多因素Logistic回归建立SNIP恶性转化的预测模型,受试者工作特征曲线(ROC)评价各独立危险因素及相关组合模型的预测能力。结果:年龄、表观扩散系数比(rADC)、流入指数(WII)2和WII3是预测SNIP恶性转化的独立危险因素,ROC曲线下面积(AUC)分别为0.767、0.862、0.727和0.704。上述危险因素组成的定量参数综合模型诊断恶性转化的AUC为0.929,敏感性为85.71%,特异性为92.77%,准确性为91.35%。结论:与评估年龄、rADC、WII等独立危险因素相比,定量参数综合模型能较好地提高SNIP与SNIP-SCC的鉴别诊断能力,为术前准确评估SNIP的恶性转化提供重要依据。

Objective:To explore the predictive value of MRI multi-parameter quantitative analysis for malignant transformation of sinonasal inverted papilloma(SNIP).Methods:Clinical and imaging data of 83 patients with SNIP and 21 patients with SNIP transformed squamous cell carcinoma(SNIP-SCC)were retrospectively reviewed.The predictive model of malignant transformation of SNIP was established by univariate analysis and multivariate Logistic regression,and the receiver operating characteristic(ROC)curve was used to evaluate the predictive ability of each independent risk factor and related combined models.Results:Age,ratio of apparent diffusion coefficient(rADC),wash in index(WII)2 and WII3 were independent risk factors for predicting malignant transformation of SNIP,and the area under ROC curve(AUC)was 0.767,0.862,0.727 and 0.704,respectively.The AUC of the quantitative parameter comprehensive model which composed of the above risk factors was 0.929,the sensitivity was 85.71%,the specificity was 92.77%,and the accuracy was 91.35%.Conclusion:Compared with the evaluation of age,rADC,and WII,the quantitative parameter comprehensive model can better improve the differential diagnosis ability of SNIP and SNIP-SCC,and provide important basis for the accurate preoperative evaluation of malignant transformation of SNIP.