基于MRI和解剖学测量的宫颈癌精准T分期诊断预测研究  

Study on the accurate T-stage diagnosis and prediction of cervical cancer based on MRI and anatomical measurement

作　　者：邹雨忻 王瑞伟 吴哲 侯文静 许杉杉 樊汝涛 王延洲 吴毅 Zou Yuxin;Wang Ruiwei;Wu Zhe;Hou Wenjing;Xu Shanshan;Fan Rutao;Wang Yanzhou;Wu Yi(Department of Digital Medicine,College of Biomedical Engineering and Medical Imaging,Army Medical University,Chongqing 400038,China;Department of Obstetrics and Gynecology,First Affiliated Hospital of Army Medical University,Chongqing 400038,China;Yu-Yue Pathology Research Center,Jinfeng Laboratory,Chongqing 401329,China;Department of Oncology,Zigong First People's Hospital,Zigong 643000,Sichuan Province,China;Department of Radiology,First Affiliated Hospital of Army Medical University,Chongqing 400038,China)

机构地区：[1]陆军军医大学生物医学工程与影像医学系数字医学教研室,重庆400038 [2]陆军军医大学第一附属医院妇产科,重庆400038 [3]金凤实验室渝粤病理科学研究中心,重庆401329 [4]自贡市第一人民医院肿瘤科,四川643000 [5]陆军军医大学第一附属医院放射科,重庆400038

出　　处：《中国临床解剖学杂志》2024年第4期382-392,共11页Chinese Journal of Clinical Anatomy

基　　金：国家自然科学基金面上项目(31971113);重庆市科技英才项目(CQYC201905037);重庆市重点研发项目(CSTB2022TIAD-KPX0181)。

摘　　要：目的确定基于磁共振成像(MRI)的宫颈癌(CC)肿瘤三维重建后的三维形态学参数测量能否精准预测其T分期诊断。方法回顾收集108例经病理证实为宫颈癌患者的术前MRI图像,将其分为T1、T2、T3和T4四组,又将T1和T2分为T1a、T1b、T2a和T2b四组。使用Amira2019软件对肿瘤、子宫、阴道、膀胱、尿道和直肠进行分割和三维重建,测量肿瘤的表面积、体积、纵径、前后径、横径、最长径、粗糙程度、质地均一程度和侵犯阴道程度。通过克鲁斯卡尔-沃利斯检验(Kruskal-Wallis test)、卡方检验(χ^(2)-test)、接收者操作特性(ROC)曲线等来评估统计学差异,并根据约登指数计算不同T分期间的临界值。结果肿瘤的表面积、体积、纵径、前后径、横径、最长径和侵犯阴道程度在T1-T4间和T1a-T2b间均有统计学差异(P<0.05),粗糙程度和质地均一程度在T1-T4间和T1a-T2b间均无统计学差异(P>0.05)。其中,T1-T4肿瘤纵径分别平均为2.82 cm、3.78 cm、4.82 cm和6.61 cm(P<0.001),纵径的临界值分别为3.17 cm、4.24 cm和6.57 cm(AUC=0.699、0.73、0.708)。T1a-T2b肿瘤纵径分别平均为2.31 cm、2.84 cm、3.63 cm和4.09 cm(P=0.008),纵径的临界值分别为2.32 cm、3.12 cm和3.94 cm(AUC=0.597、0.689、0.561)。结论MRI上宫颈癌肿瘤的形态学参数包括表面积、体积、纵径、前后径、横径、最长径和侵犯阴道程度,为预测宫颈癌患者T分期有价值的诊断因素。本研究有助于宫颈癌的精准诊断、预后评估和治疗决策。Objective To determine whether the measurement of three-dimensional morphological parameters based on magnetic resonance imaging(MRI)can accurately predict the T-stage diagnosis of cervical cancer(CC)tumors or not after three-dimensional reconstruction.Methods Preoperative MRI images from 108 patients with pathologically confirmed cervical cancer were retrospectively collected and divided into four groups:T1,T2,T3 and T4.T1 and T2 were further divided into four subgroups:T1a,T1b,T2a and T2b.The Amira2019 software was used to segment and 3D reconstruct the tumor,uterus,vagina,bladder,urethra,and rectum.The surface area,volume,longitudinal diameter,anterior posterior diameter,transverse diameter,longest diameter,roughness,texture uniformity and degree of vaginal invasion of tumor were measured.The statistical differences were evaluated by using Kruskal-Wallis test,χ^(2)-test,receiver operating characteristic(ROC)curve,etc.,and the cut-off values for different T-stages were calculated based on Youden's index.Results There were statistical differences in the surface area,volume,longitudinal diameter,anterior posterior diameter,transverse diameter,longest diameter and degree of vaginal invasion of tumor between T1-T4 and T1a-T2b(P<0.05),while there was no statistical difference in roughness and texture uniformity between T1-T4 and T1a-T2b(P<0.001).Among them,the longitudinal average diameters of T1-T4 tumors were 2.82 cm,3.78 cm,4.82 cm and 6.61 cm(P<0.001),respectively,with cut-off values of 3.17 cm,4.24 cm and 6.57 cm(AUC=0.699,0.73,0.708).The longitudinal average diameters of T1a-T2b tumors were 2.31 cm,2.84 cm,3.63 cm and 4.09 cm(P=0.008),respectively,with cut-off values of 2.32 cm,3.12 cm and 3.94 cm(AUC=0.597,0.689,0.561).Conclusions The morphological parameters of cervical cancer tumors on MRI include surface area,volume,longitudinal diameter,anterior posterior diameter,transverse diameter,longest diameter,and degree of vaginal invasion of tumor,which are valuable diagnostic factors for predicting T-stage in c

关 键 词：宫颈癌 磁共振成像 T分期诊断 三维重建 

分 类 号：R711.74[医药卫生—妇产科学] R737.33[医药卫生—临床医学] R445.2