基于图像空间重建技术的小视野DWI在直肠癌T分期中的应用价值  

Application value of reduced field of view image reconstruction using image-space sampling function DWI sequence in rectal cancer T staging

作　　者：王梓又 王冲 丁涛 刘忠啸 曾建威 张海妮 张艳春 崔莹莹 韩璐 吴鹏 孟闫凯 徐凯 WANG Zi-you;WANG Chong;DING Tao(School of Medical Imaging,Xuzhou Medical University,Jiangsu 221002,China)

机构地区：[1]徐州医科大学影像学院,江苏221002 [2]徐州医科大学附属医院影像科,江苏221002 [3]徐州医科大学附属医院病理科,江苏221002 [4]徐州医科大学附属睢宁县人民医院影像科,江苏221002 [5]飞利浦医疗保健事业部,上海200072

出　　处：《放射学实践》2024年第3期383-387,共5页Radiologic Practice

基　　金：江苏省医学会科研专项资金项目(SYH-3201150-0013(2021008));江苏省中医药科技发展计划项目(MS2021100);徐州市科学技术局重点研发计划(社会发展)(KC20159);江苏省卫生健康委老年健康科研项目(LKM2022018)。

摘　　要：目的:探讨基于图像空间重建技术的小视野DWI序列联合常规高分辨率T_(2)WI序列在直肠癌T分期中的应用价值。方法:回顾性分析2022年6月-2022年11月在本院行直肠MRI检查的患者的病例资料。所有患者均使用3.0T MR扫描仪进行检查,行常规高分辨率斜轴位T_(2)WI序列和小视野(rFOV)基于图像空间采样重建(IRIS)技术DWI(rFOV IRIS-DWI)序列扫描。所有图像均由一名从事直肠癌研究方向的影像科高级职称医生进行两轮评估。第一轮评估所有入组患者的斜轴位高分辨率T_(2)WI序列图像,一周后进行第二轮评估,同时观察斜轴位高分辨率T_(2)WI序列图像和rFOV IRIS-DWI序列图像,分别记录两次评估的磁共振T(mrT)分期结果。采用MedCalc统计软件(Version 18.2.1)进行统计学分析。两种诊断方法对病理T(pT)分期的诊断准确性采用Fisher精确概率检验。采用受试者操作特征(ROC)曲线评估两种诊断方法对pT分期的诊断效能。两种诊断方法诊断信心差值符合正态分布采用配对样本t检验,差值不符合正态分布采用配对Wilcoxon秩和检验。以P<0.05为差异具有统计学意义。结果:本研究共纳入27例直肠癌患者,年龄53~81岁,中位年龄69岁,基于rFOV IRIS-DWI联合T_(2)WI序列,分别有1例(6.7%)pT3期病例被低估为mrT1-2期、1例(8.3%)pT1-2病例被高估为mrT3期,MRI对pT1-2、pT3期的诊断准确性分别为91.7%、93.3%,两种序列的诊断准确性差异具有统计学意义(P=0.022)。T_(2)WI、T_(2)WI联合rFOV IRIS-DWI序列诊断pT3期的敏感度、特异度、ROC曲线下面积(AUC)分别为73.3%、58.3%、0.658和93.3%、91.7%、0.928,AUC差异具有统计学意义(P=0.002)。T_(2)WI、T_(2)WI联合rFOV IRIS-DWI序列对所有入组病例的诊断信心分别为2.37、2.9分,差异具有统计学意义(P=0.0002)。结论:rFOV IRIS-DWI联合高分辨率T_(2)WI序列能够明显提高直肠癌的T分期的诊断效能和诊断信心。Objective:To explore the added value of reduced field of view(rFOV)image reconstruction using image-space sampling function(IRIS)DWI(rFOV IRIS-DWI)sequence combined with T_(2)WI sequence in evaluating T staging of rectal cancer.Methods:All patients who underwent rectal MRI examination in our institution between June 2022 and August 2022 were retrospectively analyzed.All patients were performed MR scan at a 3.0T MR scanner(Ingenia Elition,Philips,Netherlands).The rFOV IRIS-DWI sequence and high-resolution oblique axial T_(2)WI sequence were scanned for analysis.All images were evaluated by a senior radiologist specialized with more than 20 years of experience in rectal cancer research by two rounds.First,the mrT staging were evaluated based on high-resolution oblique axial T_(2)WI images.Second,conducted one week later,the mrT staging were reevaluated based on rFOV IRIS-DWI combined with high-resolution oblique axial T_(2)WI images.The twice mrT staging results were recorded respectively.MedCalc statistical software(version 18.2.1)was used for data analysis.The accuracy of two diagnosis methods for pT staging were compared by Fisher's precision probability test.Receiver operating characteristic(ROC)curve was used to evaluate the diagnostic performance of two methods for pT staging.Paired sample t test was used to diagnose the confidence difference value(D-value)between the two methods if D-value was normal distribution and paired Wilcoxon S rank sum test was used if the D-value was non-normal distribution.P<0.05 indicated the difference was statistically significant.Results:Twenty-seven rectal cancer patients were enrolled in this retrospectively study.The median age of all patients was 69 years(ranged from 53 to 81 years).According to the rFOV IRIS-DWI combined with high-resolution oblique axial T_(2)WI sequence,11 patients(91.7%)were accurately diagnosed in all 12 pT1-T2 staging patients,and 14 patients(93.3%)were accurately diagnosed in all 15 pT3 staging patients.One(6.7%)pT3 patient was under-staged as mrT1-2

关 键 词：直肠癌 T分期 小视野 扩散加权成像 诊断效能 

分 类 号：R445.2[医药卫生—影像医学与核医学] R735.37[医药卫生—诊断学]