【作者】 刘洁；

【导师】 程敬亮；

【作者基本信息】 郑州大学 ， 影像医学与核医学（专业学位）， 2022， 博士

【摘要】 癌症是人体内的不成熟细胞异常增殖并分化不良而形成的恶性肿瘤,常由其起始的器官或者细胞类型来命名。宫颈癌(Cervical Cancer,CC)起源于子宫颈部,是目前包括中国在内的发展中国家妇女癌症相关死亡的第二大原因,近年来,该病的发病率随着相关筛查技术的推广和普及而逐渐提高,且发病人群趋于年轻化。人类乳头状瘤病毒(Human Papilloma Virus,HPV)感染是宫颈癌的高危因素,尤其是16、18两个亚型,其他危险因素包括低收入、性生活混乱、免疫抑制及吸烟等。确诊时宫颈癌的临床分期和肿瘤的一些组织病理学特征(如分化程度、淋巴血管间隙浸润、嗜神经浸润、基质浸润深度和淋巴结状态等)均与其治疗和预后密切相关。目前,许多影像学手段已被用于宫颈癌的诊断和预后评估,相对于超声(ultrasound,US)、CT(computed tomography)等其他影像学检查而言,磁共振成像(magnetic resonance imaging,MRI)技术在安全性、成像方式和软组织分辨率等诸多方面均具有独特优势,可以准确地显示病灶的形态、大小、边界、宫旁浸润及周围脏器侵犯,不仅为宫颈癌的早期诊断提供了丰富的信息,还为其后续治疗方案的制定及预后评估提供了重要的参考价值。目前,MRI检查已成为国际妇产科联合会(International Federation of Gynecology and Obstetrics,FIGO)推荐使用的宫颈癌分期和疗效监测的主要影像学评估手段。MRI传统成像序列主要以形态学研究为主,具有一定局限性,而自90年代发展起来的功能成像技术,实现了从细胞或分子水平对生物组织信息进行定量或半定量分析。扩散加权成像(diffusion weighted imaging,DWI)是一种通过定量参数表观扩散系数(apparent diffusion coefficient,ADC)值的测量对组织内水分子的扩散运动状态进行量化分析的定量成像技术,目前该项技术已经在宫颈癌的相关研究中得到了广泛的应用。相比于传统DWI,长可变回波链分段读出扩散加权成像(readout segmentation of long variable echo trains—diffusion weighted imaging,RESOLVE—DWI)具有图像分辨率高、显示解剖细节好、伪影少和失真度小等优点,被广泛用于头颈部、体部肿瘤的研究。T1mapping和T2mapping成像技术是近几年发展起来的用于评价组织异常特征的定量磁共振技术。T1 mapping成像技术主要通过测量图像的纵向弛豫时间(T1值)来定量分析生物组织学特征,扫描序列包括增强前及增强后T1 mapping,具有操作简单、扫描时间短等优点,以往T1mapping成像多用于心肌病变(如心肌炎、心肌纤维化及心肌淀粉样病变等)的研究,也可用于肝纤维化、慢性胰腺炎等疾病的评估;而T2mapping成像技术则通过测量图像的横向弛豫时间(T2值)定量地反映组织的含水量,其稳定性高、可重复性好,早期该技术多用于评估关节软骨病变(如膝关节损伤、椎间盘退行性变等),然而近些年来T1 mapping、T2 mapping成像技术已逐渐应用于体部肿瘤的研究。本课题主要采用T1mapping、T2mapping成像和RESOLVE—DWI扫描序列,首先,通过对宫颈癌初诊患者的血清学资料、MRI定量成像数据的收集,从以上三个技术层面对宫颈癌的临床特征(分期、血清学指标)进行相关分析;其次,通过收集患者的组织病理学资料,从不同技术层面对宫颈癌的病理学特征(如组织分型和分化、淋巴血管间隙浸润、嗜神经浸润、基质浸润深度及淋巴结状态)进行有效评估;最后,所有宫颈癌患者均接受为期6个月～3年的随访,以期分析T1 mapping、T2 mapping和RESOLVE—DWI定量成像技术对宫颈癌治疗后复发的预测价值。通过本课题研究,我们希望能够发现MRI定量成像技术在宫颈癌诊断和治疗后复发预测中的应用价值,以期为临床制定合理化治疗方案和改善患者预后提供理论支持。第一部分磁共振定量成像技术在宫预癌临床特征中的应用研究背景和目的1.比较T1 mapping、T2 mapping及RESOLVE—DWI定量成像技术对宫颈癌临床分期的诊断价值2.探讨T1mapping、T2mapping及RESOLVE—DWI定量成像技术的相关定量参数与血清鳞癌相关抗原(squamous carcinoma-associated antigen,SCC-Ag)表达水平的相关性材料与方法1、宫颈癌患者的收集:收集2018年5月至2021年4月在郑州大学第一附属医院经手术病理或宫颈活检证实的宫颈癌患者共107例,其中低分期组(ⅠB～ⅡA期)75例、高分期组(ⅡB～ⅣB期)32例;2、血清标本的收集:收集上述宫颈癌患者的血清标本107例;3、所有宫颈癌患者于治疗前均接受常规MRI、增强前后T1mapping、T2 mapping 及 RESOLVE—DWI 检查;4、在T1 mapping、T2mapping和ADC伪彩图上测量并计算肿瘤增强前纵向弛豫时间(T1pre)、增强后纵向弛豫时间(T1post)、增强后T1减少率(δT1%)、横向弛豫时间(T2)及ADC值;5、采用酶联免疫法检测宫颈癌患者血清标本中SCC-Ag的表达水平;6、采用SPSS 17.0软件进行数据的统计和分析,所有数据计量资料均采用x±s表示并进行正态性检验。符合正态分布的参数值采用独立样本t检验方法进行组间比较,以Medcalc 19.0软件绘制受试者工作特征(receiver operating characteristic,ROC)曲线,评价各参数的诊断效能,并采用DeLong检验比较曲线下面积(area under the curve,AUC)。采用Spearman相关分析各参数与临床分期、血清SCC-Ag表达水平的相关性。P<0.05为差异有统计学意义。结果1、T1 mapping、T2mapping及RESOLVE—DWI定量参数值在宫颈癌不同临床分期中的差异。①宫颈癌低分期组与高分期组对比,T2及ADC值的差异均有统计学意义(P<0.05);T2及ADC值鉴别低分期组与高分期组宫颈癌的ROC曲线AUC分别为0.840、0.687,且T2值的诊断效能优于ADC值(Z=2.237,P=0.025);②以T2=84.40 ms为最佳阈值,敏感度和特异度分别为76.01%和87.52%,约登指数为 0.635,95%置信区间为 0.756～0.904;以 ADC=0.69 × 10-3mm2/s为最佳阈值,敏感度和特异度分别为76.03%和59.41%,约登指数为0.354,95%置信区间为0.590～0.773。2、Spearman相关分析结果显示:T2值与宫颈癌临床分期呈负相关(r=-0.539,P<0.05),ADC值与宫颈癌临床分期呈负相关(r=-0.296,P<0.05);ADC值与宫颈癌患者血清SCC-Ag表达水平呈负相关(r=-0.423,P<0.001)。结论T2值及ADC值对宫颈癌临床分期的评估均有一定的诊断价值,其中T2值的诊断效能优于ADC值;而ADC值与宫颈癌患者血清SCC-Ag的表达水平具有显著的相关性。第二部分磁共振定量成像技术对宫颈癌病理学特征的诊断价值背景和目的探讨T1 mapping、T2 mapping及RESOLVE-DWI定量成像技术对宫颈癌的病理分型、分化程度、淋巴血管间隙浸润(lymphovascular space invasion,LVSI)、嗜神经浸润(perineural invasion,PNI)、基质浸润深度及淋巴结转移等病理学特征的诊断价值。材料与方法1、宫颈癌患者的收集:收集2018年5月至2021年4月在郑州大学第一附属医院经手术病理或宫颈活检证实的宫颈癌患者共107例,包括鳞癌92例、腺癌15例;宫颈鳞癌低分化组18例、中高分化组34例;LVSI阳性组31例、LVSI阴性组32例;PNI阳性组10例、PNI阴性组28例;基质浸润深度>1/2组52例、基质浸润深度≤1/2组12例;淋巴结转移阳性组20例、淋巴结转移阴性组42例;2、所有宫颈癌患者于治疗前均接受常规MRI、增强前T1mapping、增强后T1 mapping、T2 mapping 及 RESOLVE—DWI 检查;3、在T1mapping、T2mapping和ADC伪彩图上测量并计算肿瘤增强前纵向弛豫时间(T1pre)、增强后纵向弛豫时间(T1post)、增强后T1减少率(δT1%)、横向弛豫时间(T2)及ADC值;4、采用SPSS 17.0软件进行数据的统计和分析,所有数据计量资料均采用x±s表示并进行正态性检验。符合正态分布的参数值采用独立样本t检验方法比较各定量参数在宫颈癌不同组织病理学特征中的差异,通过Medcalc 19.0软件绘制ROC曲线,采用DeLong检验对比分析这些定量值对宫颈癌组织病理学特征的诊断效能,并根据约登指数确定最佳阈值。P<0.05为差异有统计学意义。结果1、宫颈鳞癌的ADC值高于腺癌,组间差异具有统计学意义(P<0.05),ROC曲线AUC为0.746,以ADC=0.70 × 10-3mm2/s作为鉴别宫颈鳞癌及腺癌的最佳阈值,敏感度、特异度分别为68.48%、80.00%,约登指数为0.485,95%置信区间为0.653～0.826;2、不同分化程度宫颈鳞癌的T1post、δT1%、T2及ADC值差异均有统计学意义(P<0.05),T1post、δT1%、T2及ADC值鉴别低分化与中高分化宫颈鳞癌的ROC 曲线 AUC 分别为 0.931、0.935、0.770、0.773,且T1post、δT1%值的诊断效能优于T2及ADC值(P<0.05),T1post与δT1%值、T2与ADC值之间的诊断效能无统计学差异(P>0.05);以T1post=626.09 ms为最佳阈值,敏感度、特异度分别为94.44%、76.47%,约登指数为0.709,95%置信区间为0.826～0.983;以8T1%=52%为最佳阈值,敏感度、特异度分别为88.89%、88.24%,约登指数为0.771,95%置信区间为0.830～0.984;以T2=83.60 ms为最佳阈值,敏感度、特异度分别为77.78%、70.59%,约登指数为0.484,95%置信区间为0.632～0.875;以ADC=0.72×10-3mm2/s为最佳阈值,敏感度、特异度分别为77.78%、67.65%,约登指数为0.454,95%置信区间为0.636～0.877;3、宫颈癌LVSI、PNI阳性组的T2值均低于LVSI、PNI阴性组,且差异均有统计学意义(P<0.05),T2值预测LVSI、PNI的ROC曲线AUC分别为0.910、0.761;以T2=84.00 ms作为预测LVSI的最佳阈值,敏感度、特异度分别为77.42%、90.62%,约登指数为0.680,95%置信区间为0.811～0.968,以T2=78.93ms作为预测PNI的最佳阈值,敏感度、特异度分别为80.00%、78.57%,约登指数为0.586,95%置信区间为0.595～0.884;基质浸润深度>1/2组的T2及ADC值均低于基质浸润深度≤1/2组,且差异具有统计学意义(P<0.05),T2及ADC值鉴别基质浸润深度的ROC曲线AUC分别为0.893、0.716,且两者的诊断效能相当(Z=1.421,P=0.155),以T2=90.54ms为最佳阈值,敏感度、特异度分别为82.69%、83.33%,约登指数为0.660,95%置信区间为0.790～0.956,以ADC=0.80X 10-3mm2/s为最佳阈值,敏感度、特异度分别为80.77%、66.67%,约登指数为0.474,95%置信区间为0.589～0.821;淋巴结转移阳性组与阴性组之间各参数值差异均无统计学意义(P>0.05)。结论T1 mapping、T2mapping及RESOLVE—DWI定量成像技术在宫颈癌组织病理学特征的评估方面具有不同的诊断价值。第三部分磁共振定量成像技术对宫颈癌治疗后复发的预测价值背景和目的探讨T1 mapping、T2mapping及RESOLVE—DWI定量成像技术对宫颈癌治疗后复发的预测价值。材料与方法1、宫颈癌患者的收集:收集2018年5月至2021年4月在郑州大学第一附属医院经手术病理或宫颈活检证实的宫颈癌患者共107例,其中包括手术组77例(复发组13例、非复发组64例)、非手术组30例(复发组10例、非复发组20例);2、对手术或同步放化疗治疗后的宫颈癌患者进行为期6个月～3年的随访;3、所有宫颈癌患者均于治疗前接受常规MRI、增强前后T1mapping、T2 mapping 及 RESOLVE—DWI 检查;4、在T1 mapping、T2mapping和ADC伪彩图上测量并计算肿瘤的增强前纵向弛豫时间(T1pre)、增强后纵向弛豫时间(T1post)、增强后T1减少率(δT1%)、横向弛豫时间(T2)及ADC值;5、采用SPSS 17.0软件进行数据的统计和分析,所有数据计量资料均采用x±s表示并进行正态性检验。符合正态分布的数据采用独立样本t检验方法进行组间比较,以Medcalc 19.0软件绘制ROC曲线,评价相关参数对宫颈癌治疗后复发的预测效能,并根据约登指数确定最佳阈值。采用Logistic回归分析影响宫颈癌患者治疗后复发的因素。P<0.05为差异有统计学意义。结果1、107例宫颈癌患者治疗后复发23例,总复发率21.5%,其中手术组77例,复发13例,复发率16.9%,非手术组30例,复发10例,复发率33.3%;2、T1 mapping、T2mapping及RESOLVE—DWI定量参数值在预测宫颈癌手术治疗后复发组与非复发组间的差异。①手术组中复发组与非复发组对比,T1pre值的差异有统计学意义(P<0.05);②T1pre值预测宫颈癌术后复发的ROC曲线AUC为0.742,以T1pre=1480.19 ms作为最佳阈值,敏感度、特异度分别为76.92%、70.31%,约登指数为0.472,95%置信区间为0.630～0.825;3、T1 mapping、T2 mapping及RESOLVE—DWI定量参数值在预测宫颈癌非手术治疗后复发组与非复发组间的差异。①非手术组中复发组与非复发组对比,T1pre值的差异有统计学意义(P<0.05);②T1pre值预测宫颈癌非手术治疗后复发的ROC曲线AUC为0.780,以T1pre=1494.00 ms作为最佳阈值,敏感度、特异度分别为80.00%、75.00%,约登指数为0.550,95%置信区间为0.592～0.910。4、Logistic回归分析显示T1pre值是宫颈癌患者治疗后复发的预测因素。结论治疗前T1 mapping成像的T1pre值对宫颈癌治疗后复发有一定的预测价值。更多还原

【Abstract】 Cancer is a malignant tumor formed by abnormal proliferation and poor differentiation of immature cells in the human body.It is often named by its initial organ or cell type.Cervical cancer(CC),which originated from the cervix,is the second leading cause of cancer related death among women in developing countries including China.The incidence rate of this disease has gradually improved with the promotion and popularization of screening technology in recent years,and the incidence of CC tends to be younger.Human papillomavirus(HPV)infection is a high-risk factor for CC,especially subtypes 16 and 18.Other risk factors include low income,sexual life disorder,immunosuppression and smoking.The clinical stage and some pathological features of the tumor(such as degree of differentiation,lymphovascular space invasion,perineural invasion,depth of matrix infiltration,lymph node status,etc.)are closely related to its treatment and prognosis.At present,many imaging methods have been used in the diagnosis and prognosis of CC.Compared with other imaging examinations such as ultrasound(US),computed tomography(CT),magnetic resonance imaging(MRI)technology has unique advantages in safety,imaging mode and soft tissue resolution.It can accurately display the morphology,size,boundary,parametrial infiltration and surrounding organ invasion of the lesion,which is not only of great value in the early diagnosis of CC,but also useful for the formulation of treatment plan and prognosis evaluation.The International Federation of gynecology and Obstetrics(FIGO)suggested that MRI should be used as the main imaging examination method for the staging and efficacy detection of CC.The traditional MRI imaging sequence mainly focuses on morphological research,which has some limitations.The functional imaging technology developed since the 1990s realizes the quantitative or semi quantitative analysis of biological tissue information from the cellular or molecular level.Diffusion weighted imaging(DWI)is a quantitative imaging technique that reflects the diffusion movement of water molecules in living tissues.It quantitatively analyzes the diffusion movement of water molecules in diseased tissues by measuring the apparent diffusion coefficient(ADC).At present,DWI has been widely used in the study of clinical staging and pathological characteristics of CC.Compared with the traditional DWI sequence,readout segmentation of long variable echo trains-diffusion weighted imaging(RESOLVE-DWI)has the advantages of high image resolution,good display of anatomical details,less artifacts and low distortion.It has been widely used in the research of head,neck and body tumors.T1 mapping and T2 mapping are novel quantitative MRI techniques developed in recent years to evaluate the characteristics of tissue abnormalities.T1 mapping quantitatively analyzes the biological histological characteristics by measuring the longitudinal relaxation time(T1 value)of the image.The scanning sequence includes T1 mapping before and after enhancement.It has the advantages of simple operation and short scanning time.T1 mapping was mostly used to evaluate myocardial lesions(such as myocarditis,myocardial fibrosis and myocardial amyloid),as well as liver fibrosis,chronic pancreatitis and other diseases in past years.T2 mapping quantitatively reflects the tissue water content by measuring the transverse relaxation time(T2 value)of the image,which has high stability and repeatability.In the past,T2 mapping was mostly used to evaluate articular cartilage lesions(such as knee injury and intervertebral disc degeneration).In recent years,T1 mapping and T2 mapping have been gradually applied to the study of body tumors.Therefore,this subject mainly used T1 mapping,T2 mapping and RESOLVEDWI scanning sequence.Firstly,through the collection of serological data and MRI quantitative imaging data of newly diagnosed patients with CC,the clinical characteristics of CC(stage,serological index)were analyzed from the above three technical levels;secondly,by collecting the histopathological data of patients,the pathological characteristics of CC(such as tissue type,degree of differentiation,lymphovascular space invasion,perineural invasion,matrix infiltration depth and lymph node status)were effectively evaluated from different technical levels;finally,all patients with CC were followed up for 6 months to 3 years in order to analyze the predictive value of T1 mapping,T2 mapping and RESOLVE-DWI in the recurrence of CC after treatment.We hope to find the application value of MRI quantitative imaging technology in the diagnosis and recurrence prediction of CC through this research,in order to provide theoretical support for the rational treatment and prognosis improvement of CC.Part Ⅰ The application of quantitative MRI in clinical features of cervical cancerBackground and Purpose1.To compare the diagnostic value of T1 mapping,T2 mapping and RESOLVEDWI in differentiating clinical stage of cervical cancer(CC);2.To clarify the correlation between the quantitative parameters of T1 mapping,T2 mapping,RESOLVE-DWI and the expression level of squamous carcinoma-associated antigen(SCC-Ag)in serum of patients with CC.Materials and Methods1.Collection of patients with CC:a total of 107 patients with CC confirmed by surgery pathology or cervical biopsy in the First Affiliated Hospital of Zhengzhou University from May 2018 to April 2021 were collected,including 75 cases in the low-stage group(stage ⅠB～ⅡA)and 32 cases in the high-stage group(stage ⅡB～ⅣB);2.Collection of serum samples:107 serum samples from the above CC patients were collected;3.All patients with CC underwent routine MRI,T1 mapping before and after enhancement,T2 mapping and RESOLVE-DWI;4.The longitudinal relaxation time before enhancement(T1pre),longitudinal relaxation time after enhancement(T1post),T1 reduction rate after enhancement(δ T1%),transverse relaxation time(T2)and ADC values were measured and calculated on T1 mapping,T2 mapping and ADC pseudo-color images automatically generated by Siemens post-processing workstation;5.The expression of SCC-Ag in serum samples of patients with CC was detected by enzyme-linked immunosorbent assay;6.SPSS 17.0 software was used for data statistics and analysis.All data were expressed as the means±standard error of the mean and tested for normality test.The parameter values conforming to the normal distribution were compared between groups by independent sample t-test,the receiver operating characteristic(ROC)curve was constructed by medcalc 19.0 software,the diagnostic performances of each parameter were evaluated,and the area under the curve(AUC)was compared by Delong test.Spearman correlation was used to analyze the correlation between parameters and clinical stage and serum SCC-Ag expression level.P<0.05 was statistically significant.Results1.Differences of quantitative parameters of T1 mapping,T2 mapping and RESOLVE-DWI in different clinical stages of CC.① There were significant differences in T2 value and ADC value between low-stage group and high-stage group(P<0.05);The AUC of ROC curve of T2 and ADC in differentiating low-stage group and high-stage group were 0.840 and 0.687 respectively,and the diagnostic efficiency of T2 was better than ADC(Z=2.237,P=0.025);②Taking T2=84.40 ms as the optimal threshold,the sensitivity and specificity were 76.01%and 87.52%respectively,the youden index was 0.635,and 95%confidence interval was 0.756～0.904;taking ADC=0.69×10-3mm2/s as the best threshold,the sensitivity and specificity were 76.03%and 59.41%respectively,the youden index was 0.354,and the 95%confidence interval was 0.590～0.773.2.Spearman correlation analysis showed that T2 value was negatively correlated with clinical stage of CC(r=-0.539,P<0.05),and ADC value was negatively correlated with clinical stage of CC(r=-0.296,P<0.05);ADC value was negatively correlated with serum SCC-Ag expression in patients with CC(r=-0.423,P<0.001).ConclusionsBoth T2 and ADC values are helpful parameter to evaluate the clinical stage of CC,while the diagnostic efficiency of T2 value is better than ADC value;ADC value is significantly correlated with the expression level of serum SCC-Ag in patients with CC.Part II The diagnostic value of quantitative MRI in pathological features of cervical cancerBackground and PurposeTo investigate the diagnostic value of T1 mapping,T2 mapping and RESOLVEDWI in the pathological classification,degree of differentiation,lymphovascular space invasion(LVSI),perineural invasion(PNI),depth of stromal invasion and lymph node metastasis of cervical cancer(CC).Materials and Methods1.Collection of patients with CC:a total of 107 patients with CC confirmed by surgery pathology or cervical biopsy in the First Affiliated Hospital of Zhengzhou University from May 2018 to April 2021 were collected,including 92 cases of squamous cell carcinoma and 15 cases of adenocarcinoma;18 cases in poorly differentiated group and 34 cases in well/moderately differentiated group of cervical squamous cell carcinoma;31 cases in LVSI-positive group and 32 cases in LVSI-negative group;10 cases in PNI-positive group and 28 cases in PNI-negative group;52 cases in the group with matrix infiltration depth>1/2 and 12 cases in the group with matrix infiltration depth ≤ 1/2;20 cases with positive lymph node metastasis and 42 cases with negative lymph node metastasis;2.All patients with CC underwent routine MRI,T1 mapping before and after enhancement,T2 mapping and RESOLVE-DWI;3.The longitudinal relaxation time before enhancement(T1pre),longitudinal relaxation time after enhancement(T1post),T1 reduction rate after enhancement(δT1%),transverse relaxation time(T2)and ADC values were measured and calculated on T1 mapping,T2 mapping and ADC pseudo-color images automatically generated by Siemens post-processing workstation;4.SPSS 17.0 software was used for data statistics and analysis.All data were expressed as the means±standard error of the mean and tested for normality test.For the parameter values conforming to the normal distribution,the quantitative parameters in different pathological characteristics of CC were compared by independent sample t-test,the receiver operating characteristic(ROC)curve was constructed by medcalc 19.0 software,and the diagnostic performances of these quantitative values on the pathological characteristics of CC were compared and analyzed by Delong test,The optimal thresholds were determined according to the youden index.P<0.05 was statistically significant.Results1.The ADC value of cervical squamous cell carcinoma was significantly higher than that of adenocarcinoma(P<0.05),and the area under the curve(AUC)of ROC curve was 0.746,taking ADC=0.70×10-3mm2/s as the best threshold to distinguish cervical squamous cell carcinoma and adenocarcinoma,the sensitivity and specificity were 68.48%and 80.00%respectively,the youden index was 0.485,and the 95%confidence interval was 0.653～0.826;2.There were significant differences in T1post,δT1%,T2 and ADC values of cervical squamous cell carcinoma with different differentiation degrees(P<0.05).The AUC of ROC curve of T1post,δT1%,T2 and ADC values in differentiating poorly differentiated and well/moderately differentiated squamous cell carcinoma were 0.931,0.935,0.770 and 0.773 respectively,and the diagnostic efficiencies of T1post and δT1%values were better than T2 and ADC values(P<0.05),there was no significant difference in diagnostic efficacy between T1post and δT1%,T2 and ADC(P>0.05);taking T1post=626.09 ms as the best threshold,the sensitivity and specificity were 94.44%and 76.47%respectively,the youden index was 0.709,and the 95%confidence interval was 0.826～0.983;taking δT1%=52%as the best threshold,the sensitivity and specificity were 88.89%and 88.24%respectively,the youden index was 0.771,and the 95%confidence interval was 0.830～0.984;taking T2=83.60 ms as the best threshold,the sensitivity and specificity were 77.78%and 70.59%respectively,the youden index was 0.484,and the 95%confidence interval was 0.632～0.875;taking ADC=0.72×10-3mm2/s as the best threshold,the sensitivity and specificity are 77.78%and 67.65%respectively,the youden index is 0.454,and the 95%confidence interval is 0.636～0.877;3.The T2 values of LVSI and PNI positive group were lower than those of LVSI and PNI negative group,and the difference was statistically significant(P<0.05).The AUC of ROC curve predicted by T2 value was 0.910 and 0.761,respectively;taking T2=84.00 ms as the best threshold for predicting LVSI,the sensitivity and specificity were 77.42%and 90.62%respectively,the youden index was 0.680,and the 95%confidence interval was 0.811～0.968;taking T2=78.93 ms as the best threshold for predicting PNI,the sensitivity and specificity were 80.00%and 78.57%respectively,the youden index was 0.586,and the 95%confidence interval was 0.595～0.884;the values of T2 and ADC in the group with matrix infiltration depth>1/2 were lower than those in the group with matrix infiltration depth ≤1/2,and the difference was statistically significant(P<0.05).The AUC of ROC curve for T2 and ADC values to distinguish the depth of matrix infiltration were 0.893 and 0.716 respectively,and their diagnostic efficiency was the same(Z=1.421,P=0.155),taking T2=90.54 ms as the optimal threshold,the sensitivity and specificity were 82.69%and 83.33%respectively,the youden index was 0.660,and the 95%confidence interval was 0.790～0.956,taking ADC=0.80×10-3mm2/s as the best threshold,the sensitivity and specificity were 80.77%and 66.67%respectively,the youden index was 0.474,and the 95%confidence interval was 0.589～0.821.There was no significant difference between the lymph node metastasis positive group and the negative group(P>0.05).ConclusionsThe quantitative parameters of T1 mapping,T2 mapping and RESOLVE-DWI could have different diagnostic values in the evaluation of histopathological features of cervical cancer.Part Ⅲ The value of quantitative MRI in predicting recurrence of cervical cancer after treatmentBackground and PurposeTo investigate the value of T1 mapping,T2 mapping and RESOLVE-DWI in predicting the recurrence of cervical cancer(CC)after treatment.Materials and Methods1.Collection of CC patients:a total of 107 patients with CC confirmed by surgery pathology or cervical biopsy in the First Affiliated Hospital of Zhengzhou University from May 2018 to April 2021 were collected,including 77 cases in the operation group(13 cases in the recurrence group and 64 cases in the non recurrence group)and 30 cases in the non operation group(10 cases in the recurrence group and 20 cases in the non recurrence group);2.Patients with CC after surgery or concurrent chemoradiotherapy were followed up for 6 months to 3 years;3.All patients with CC underwent routine MRI,T1 mapping before and after enhancement,T2 mapping and RESOLVE-DWI;4.The longitudinal relaxation time before enhancement(T1pre),longitudinal relaxation time after enhancement(T1post),T1 reduction rate after enhancement(δ T1%),transverse relaxation time(T2)and ADC values were measured and calculated on T1 mapping,T2 mapping and ADC pseudo-color images automatically generated by Siemens post-processing workstation;5.SPSS 17.0 software was used for data statistics and analysis.All data were expressed as the means±standard error of the mean and tested for normality test.The parameter values conforming to the normal distribution were compared between groups by independent sample t-test method.The receiver operating characteristic(ROC)curve was constructed by medcalc 19.0 software to evaluate the prediction performances of relevant parameters on the recurrence of CC after treatment,and the optimal thresholds were determined according to the youden index.Logistic regression analysis was used to analyze the factors influencing the recurrence of CC patients after treatment.P<0.05 was statistically significant.Results1.Among 107 patients with CC,23 cases recurred after treatment,the total recurrence rate of 21.5%.13 cases recurred in the operation group(n=77),the recurrence rate of 16.9%;10 cases recurred in the non operation group(n=30),the recurrence rate of 33.3%;2.The quantitative parameters of T1 mapping,T2 mapping and RESOLVE-DWI were used to predict the difference between recurrent group and non recurrent group after surgery of CC.① There was significant difference in T1pre value between recurrent group and non recurrent group(P<0.05);② The area under the curve(AUC)of ROC curve for predicting postoperative recurrence of CC by T1pre value was 0.742,taking T1pre=1480.19 ms as the best threshold,the sensitivity and specificity were 76.92%and 70.31%respectively,the youden index was 0.472,and the 95%confidence interval was 0.630～0.825;3.The difference of T1 mapping,T2 mapping and RESOLVE-DWI quantitative parameters in predicting the recurrence group and non recurrence group after non-surgical treatment of CC.① There was significant difference in T1pre value between recurrent group and non recurrent group in non-surgical group(P<0.05);②The AUC of ROC curve for predicting the recurrence of CC after non-surgical treatment with T1pre value was 0.780,taking T1pre=1494.00 ms as the best threshold,the sensitivity and specificity were 80.00%and 75.00%respectively,the youden index was 0.550 and the 95%confidence interval was 0.592～0.910.4.Logistic regression analysis showed that T1pre value was a predictor of recurrence in patients with CC after treatment.ConclusionsThe T1pre value of T1 mapping imaging before treatment is of significance for predicting the recurrence of CC after treatment.更多还原