摘要
目的:探讨合成MRI定量弛豫参数对于腮腺常见肿瘤的鉴别诊断.方法:选择2022 年5 月至2023 年1月郑州大学第一附属医院收治的腮腺肿瘤患者59 例,均于治疗前行常规MRI和合成MRI,测量肿瘤全域弛豫参数[纵向弛豫时间(T1)、横向弛豫时间(T2)和质子密度(PD)].比较腮腺多形性腺瘤组、Warthin瘤组和恶性肿瘤组T1、T2 和PD.应用Bayes判别分析法对3 种腮腺肿瘤进行分类.结果:腮腺多形性腺瘤、Warthin瘤和恶性肿瘤组T1、T2 和PD的差异均有统计学意义(P均<0.001).Warthin瘤组的T1 和PD小于腮腺恶性肿瘤和多形性腺瘤组(P<0.05);腮腺多形性腺瘤组的T2 大于恶性肿瘤和Warthin瘤组(P<0.05).采用Bayes判别分析建立的诊断模型自身验证的准确率为61.0%,交叉验证的准确率为 57.6%.结论:合成MRI定量弛豫参数对腮腺多形性腺瘤、Warthin瘤和恶性肿瘤的鉴别诊断有一定价值.
Abstract
Aim:To investigate the differential diagnostic value of quantitative relaxation parameters derived from syn-thetic MRI for common tumors of the parotid gland.Methods:A total of 59 patients with parotid gland tumors admitted to the First Affiliated Hospital of Zhengzhou University from May 2022 to January 2023 were selected.All patients underwent conventional MRI and synthetic MRI before treatment,and whole-tumor relaxation parameters including longitudinal relaxa-tion time(T1),transverse relaxation time(T2),and proton density(PD)of the tumors were measured.T1,T2,and PD were compared among the pleomorphic adenoma group,Warthin tumor group,and malignant tumor group.Bayes discriminant a-nalysis was applied to classify the 3 types of parotid gland tumors.Results:Significant differences in T1,T2,and PD were observed among the groups of pleomorphic adenoma,Warthin tumor,and malignant tumors(all P<0.001).T1 and PD of Warthin tumors were lower than those of malignant tumors and pleomorphic adenomas(P<0.05);T2 of pleomorphic adeno-mas were higher than those of malignant tumors and Warthin tumors(P<0.05).The accuracy of the diagnostic model es-tablished by Bayes discriminant analysis for self-validation was 61.0%,and that for cross-validation was 57.6%.Conclu-sion:Quantitative relaxation parameters based on synthetic MRI are helpful for the classification and identification of parotid pleomorphic adenomas,Warthin tumors,and malignant tumors.
基金项目
工信部公共服务平台项目(2020-0103-3-1)
国家科技期刊平台
NSTL
万方数据