1、98中国CT和MRI杂志2023年02月 第21卷 第02期 总第160期【第一作者】田甜，女，主治医师，主要研究方向：放射影像诊断。E-mail：【通讯作者】云昊，男，主治医师，主要研究方向：放射影像诊断。E-mail：论 著Relationship between Signal Intensity Histogram of Global Diffusion Coefficient and Pathological Stage in Breast Cancer and Its Predictive Value for RecurrenceTIAN Tian,YUN Hao*,WANG Sha
2、n,LUO Fa-fu.Department of Radiology,Wenchang Peoples Hospital,Wenchang 571399,Hainan Province,ChinaABSTRACTObjective To explore the relationship between the global apparent diffusion coefficient(ADC)signal intensity histogram and pathological stage of breast cancer and its predictive value for recur
3、rence.Methods The clinical,pathological and imaging data of 90 patients with invasive ductal carcinoma treated in our hospital from June 2016 to July 2020 were retrospectively collected,all patients underwent MRI examination.Apply Siemens workstation to rebuild ADC image,Extract 3D region of interes
4、t(ROI)ADC minimum value(ADCmin),ADC median(ADCmedian),ADC mean(ADCmean),ADC mode(ADCmode),ADC standard deviation(ADCstd),ADC maximum value(ADCmax),10th percentile(ADC10%),25th percentile(ADC25%),50th percentile(ADC50%),75th percentile(ADC75%),90th percentile(ADC90%),skewness coefficient and kurtosis
5、 coefficient and other parameters.The patients were divided into groups according to tumor pathological type and postoperative recurrence,the differences between the groups were compared,and the diagnostic and predictive efficacy of ADC histogram parameters was evaluated by drawing receiver operatin
6、g characteristic curve(ROC).Results ADC25%,ADC50%and ADC95%in stage T1-T2 group were significantly lower than those in stage T3-T4 group,and the kurtosis coefficient was significantly higher than that in stage T3-T4 group,with statistical significance(P0.05).In ADC histogram,the diagnostic efficacy
7、of ADC95%in distinguishing different pathological stages of breast cancer was the best.When the best diagnostic threshold was 1.36310-3mm2/s,the area under the curve was 0.869,and the corresponding Youden index,sensitivity and specificity were 0.741,88.50%and 85.60%,respectively.Compared with the no
8、n-recurrent group,the ADC10%,ADC25%,ADC90%and kurtosis coefficient of recurrent breast cancer patients were significantly decreased,with statistical significance(P0.05),and all of them had certain predictive value for the recurrence of breast cancer patients.Compared with the single prediction index
9、,the combined prediction had higher accuracy.Conclusions ADC25%,ADC50%,ADC95%and kurtosis coefficient all have certain diagnostic value for the pathological staging of breast cancer,and ADC10%,ADC25%,ADC90%and kurtosis coefficient have certain application value in predicting the recurrence of breast
10、 cancer patients.Keywords:Breast Cancer;Apparent Diffusion Coefficient;Diffusion Weighted Imaging;Pathological Stage;Recrudescence乳腺癌是女性常发的恶性肿瘤，具有高度异质性，其预后与病理分期、组织级别及淋巴结转移等因素有关1-2。其中病理分期是重要的预后因素，术前准确判断乳腺癌病理分期对临床治疗及预后评估具有重要意义。近年来，扩散加权成像(diffusion-weighted imaging，DWI)被广泛应用于肿瘤病理分级及诊断3，其中表观扩散系数(apparen
11、t diffusion coefficient，ADC)可定量评估组织水分子扩散受限特征，与局部ADC值相比，肿瘤全域的ADC值在反映肿瘤整体异质性方面准确性更高，并可避免局部勾画感兴趣区(region of interest，ROI)带来的抽样误差4-5。直方图分析是纹理分析的重要组成部分，可通过分析像素分布获得多个直方图参数，从而体量化肿瘤的异质性6。目前，直方图分析已广泛应用于乳腺癌分子分型7、病理类型8及预后9方面的研究，但关于ADC直方图鉴别乳腺癌病理分期的研究还鲜有报道，因此，本研究旨在探讨基于肿瘤全域的ADC信号强度直方图与乳腺癌病理分期的相关性及其对患者术后复发的预测价值。1
12、资料与方法1.1 一般资料 回顾性收集2016年6月至2020年7月于我院接受治疗的90例浸润性导管癌患者的临床、病理及影像学资料，患者均为女性，年龄2673岁，中位年龄49岁。纳入标准：术前乳腺磁共振成像(magnetic resonance imaging，MRI)检查提示肿块型病变，术后确诊为乳腺癌；MRI的DWI及ADC图像质量能确切显示病灶；无核磁共振检查禁忌；均为单侧单发病变。排除标准：已确诊为浸润性导管癌并接受过手术或新辅助治疗；3个月内接受过乳腺穿刺活检；病灶最大直径1cm者；临床资料不全、DWI图像质量欠佳者。1.2 MRI检查 所有患者均在MAGNETOM Verio 3.
13、0 T MR(德国西门子)扫描仪上进行MRI检查，患者取俯卧位，双乳自然悬垂于8通道相控表面线圈内，扫描T1WI、DWI及T2WI动态增强序列，动态增强前先扫描蒙片，然后静脉团注Gd-DTPA，剂量为0.1 mmol/乳腺癌肿瘤全域的表观扩散系数信号强度直方图与病理分期的关系及对复发的预测价值田 甜 云 昊*王 山罗发福文昌市人民医院放射科(海南 文昌 571399)【摘要】目的 探讨乳腺癌肿瘤全域的表观扩散系数(ADC)信号强度直方图与病理分期的关系及对复发的预测价值。方法 回顾性收集2016年6月-2020年7月于我院接受治疗的90例浸润性导管癌患者的临床、病理及影像学资料，患者均接受MR
14、I检查，应用西门子相应工作站重建ADC图，提取3D 感兴趣区(ROI)ADC最小值(ADCmin)、ADC 中位数(ADCmedian)、ADC平均值(ADCmean)、ADC众数(ADCmode)、ADC标准差(ADCstd)、ADC 最大值(ADCmax)、第10百分位数(ADC10%)、第25百分位数(ADC25%)、第50百分位数(ADC50%)、第75百分位数(ADC75%)、第90百分位数(ADC90%)、偏度系数及峰度系数等参数，根据肿瘤病理类型及术后复发情况进行分组，对比组间差异，通过绘制受试者工作特征曲线(ROC)评估ADC直方图参数的诊断和预测效能。结果 乳腺癌T1T2期组
15、ADC25%、ADC50%、ADC95%均显著低于T3T4期组，而峰度系数则显著高于T3T4期组，差异具有统计学意义(P0.05)。ADC直方图中，ADC95%鉴别乳腺癌不同病理分期诊断效能最佳，以1.36310-3mm2/s作为最佳诊断阈值时，其曲线下面积为0.869，对应的Youden指数、敏感度和特异度分别为0.741、88.50%和85.60%。与非复发组比较，复发乳腺癌患者ADC10%、ADC25%、ADC90%及峰度系数均显著降低，差异具有统计学意义(P0.05)，且均对乳腺癌患者复发具有一定的预测价值，相比于单一预测指标，联合预测具有更高的准确性。结论 ADC25%、ADC50%
16、、ADC95%及峰度系数均对乳腺癌病理分期有一定的诊断价值，ADC10%、ADC25%、ADC90%及峰度系数在预测乳腺癌患者复发方面具有一定的应用价值。【关键词】乳腺癌；表观扩散系数；扩散加权成像；病理分期；复发【中图分类号】R737.9【文献标识码】A DOI:10.3969/j.issn.1672-5131.2023.02.034 99CHINESE JOURNAL OF CT AND MRI,FEB.2023,Vol.21,No.02 Total No.160kg，速率为2.5mL/s，总量15mL，经肘静脉建立静脉通道。平均扫序列扫描参数：T1WI：TR 764ms，TE 8.1ms，层厚4mm，层数35层，层间距1，FOV 340 mm 340mm144mm，激励次数2；T2WI：TR 3200ms，TE 101ms，层厚4mm，层数35层，层间距1mm，FOV 340mm 340mm144mm，激励次数1；DWI采用横断面脂肪抑制平面回波序列，TR 6600ms，TE 89ms，层厚5mm，层间距1mm，b 值0s/mm2、800s/mm2，FOV 300mm 166mm
