多参数磁共振成像联合表观弥散系数直方图分析预测膀胱癌的病理分化  

Multiparametric MRI Combined with Apparent Diffusion Coefficient Histogram Analysis for Assessing Variant Histology in Urothelial Carcinoma

作　　者：孔令敏 凌坚[2] 蔡迁 温志华[1] 郭燕[1] 王焕军[1] KONG Ling-min;LING Jian;CAI Qian;WEN Zhi-hua;GUO Yan;WANG Huan-jun(Department of Radiology,The First Affiliated Hospital,Sun Yat-sen University,Guangzhou 510080,China;De-partment of Radiology,The East Division of the First Affiliated Hospital,Sun Yat-Sen University,Guangzhou 510700,China)

机构地区：[1]中山大学附属第一医院放射科,广东广州510080 [2]中山大学附属第一医院东院放射科,广东广州510700

出　　处：《中山大学学报（医学科学版）》2023年第6期1008-1015,共8页Journal of Sun Yat-Sen University:Medical Sciences

基　　金：国家自然科学基金(82071989,82371911,82372075)。

摘　　要：【目的】探讨多参数磁共振成像联合表观弥散系数(ADC)直方图分析应用于膀胱癌合并其他病理分化类型的预测价值。【方法】回顾性收集并分析2015年3月至2023年3月就诊于中山大学附属第一医院并经手术病理证实为膀胱尿路上皮癌患者的临床及影像学资料。根据术后病理结果,将患者分为有合并其他病理分化(如鳞状分化、腺样分化等)组及无其他病理分化组。运用倾向性评分匹配的方法,将无分化组患者按年龄及性别与有分化组患者进行两组间1∶1最邻近匹配,两组均入组49例。使用ITK-SNAP软件在ADC对患者病灶全层勾画并将其导入开源软件Pyradiomics提取直方图特征。比较两组间膀胱癌患者的临床、病理及影像参数差异,计算受试者工作特征曲线(ROC)的曲线下面积(AUC)评估各参数鉴别膀胱癌有无分化的诊断效能。采用多因素logistic回归筛选独立危险因素,以此构建临床模型。【结果】病灶形态、Ⅵ-RADS评分、强化特点及周围是否有可疑淋巴结肿大在两组间差异有统计学意义(P<0.05)。有分化组ADC均数、第10、25、75和90百分位数均小于未分化组(P<0.05)。多因素logistic回归分析发现,病灶强化特点、ADC平均数及第25,75百分位数为独立预测因子(P<0.05),构建联合模型,其诊断效能最优[AUC(95%CI)=0.91(0.83,0.96)]。【结论】多参数磁共振成像联合ADC直方图分析可准确预测膀胱癌患者是否合并其他病理分化,可协助临床制定更个体化的治疗方案。【Objective】To investigate the feasibility of multiparametric MRI(mpMRI)combined with histogram analy‐sis of apparent diffusion coefficient(ADC)in the assessment of patients with variant histology(VH)of urothelial carcino‐ma(UC).【Method】We retrospectively analyzed the data of patients pathologically diagnosed with UC who underwent mpMRI in the First Affiliated Hospital of Sun Yat-sen University between March 2015 and March 2023.The patients were divided into VH group(urothelial carcinoma mixed with other histologies)and non-VH group(pure urothelial carcinoma)according to pathological results.We performed propensity score 1:1 nearest-neighbor matching on the two groups based on age and gender and 49 patients were included in each group.The regions of interest(ROIs)of the whole tumor were de‐lineated manually by using ITK-SNAP software and Pyradiomics was applied to extract ADC histogram parameters.We compared the clinicopathological data,MRI morphological features and ADC histogram parameters between the groups.Multivariate logistic regression was used to identify independent risk factors and construct the prediction model.Receiver operating characteristic(ROC)curve analyses were performed to evaluate the diagnostic performance of these parameters for determining VH of UC.【Results】MRI morphological features including the lesion shape,vesical imaging-reporting and data system(Ⅵ-RADS)score,enhancement pattern and suspicious lymph node metastasis were markedly different between the two groups(all P<0.05).ADC mean,ADC median,ADC25th,ADC75th,ADC10th and ADC90th were significantly lower in patients with VH than those in non-VH group(all P<0.05).Multivariate logistic regression analysis showed enhance‐ment pattern,ADC25th,ADC75th and ADC mean were independent predictors(P<0.05).The combined model yielded the best predictive performance,with an area under the ROC curve(AUC)of 0.91(95%CI:0.83-0.96).【Conclusions】MpMRI combined with whole-tumor histogram analysis of ADC can serve as a reliable method

关 键 词：膀胱肿瘤 尿路上皮 磁共振成像 直方图 分化 

分 类 号：R445.2[医药卫生—影像医学与核医学]