en
×

分享给微信好友或者朋友圈

使用微信“扫一扫”功能。
通讯作者:

何健,E-mail:hjxueren@126.com

中图分类号:R735.9

文献标识码:A

文章编号:1007-4368(2022)10-1464-08

DOI:10.7655/NYDXBNS20221019

参考文献 1
KARDOSH A,LICHTENSZTAJN D,GUBENS M,et al.Long ⁃ term survivors of pancreatic cancer:a California population⁃based study[J].Pancreas,2018,47(8):958-966
参考文献 2
NEESSE A,BAUER C A,ÖHLUND D,et al.Stromal biol⁃ ogy and therapy in pancreatic cancer:ready for clinical translation?[J].Gut,2019,68(1):159-171
参考文献 3
WASIF N,KO C Y,FARRELL J,et al.Impact of tumor grade on prognosis in pancreatic cancer:should we in⁃ clude grade in AJCC staging?[J].Ann Surg Oncol,2010,17(9):2312-2320
参考文献 4
JEEKEL H.Prognostic factors following curative resec⁃ tion for pancreatic adenocarcinoma[J].Ann Surg,2004,240(2):384
参考文献 5
WINTER J,CAMERON J,CAMPBELL K.1423 pancreat⁃ icoduodenectomies for pancreatic cancer:a single⁃institu⁃ tion experience[J].J Gastrointest Surg,2006,10(9):1199-1121
参考文献 6
MACÍAS N,SAYAGUÉS J M,ESTEBAN C,et al.Histo⁃ logic tumor grade and preoperative bilary drainage are the unique independent prognostic factors of survival in pan⁃ creatic ductal adenocarcinoma patients after pancreatico⁃ duodenectomy[J].J Clin Gastroenterol,2018,52(2):e11-e17
参考文献 7
CASSINOTTO C,CHONG J,ZOGOPOULOS G,et al.Re⁃ sectable pancreatic adenocarcinoma:role of CT quantita⁃ tive imaging biomarkers for predicting pathology and pa⁃ tient outcomes[J].Eur J Radiol,2017,90:152-158
参考文献 8
NURMI A,MUSTONEN H,PARVIAINEN H,et al.Neo⁃ adjuvant therapy offers longer survival than upfront sur⁃ gery for poorly differentiated and higher stage pancreatic cancer[J].Acta Oncol,2018,57(6):799-806
参考文献 9
GERLINGER M,ROWAN A J,HORSWELL S,et al.In⁃ tratumor heterogeneity and branched evolution revealed by multiregion sequencing[J].N Engl J Med,2012,366(10):883-892
参考文献 10
YEH R,DERCLE L,GARG I,et al.The role of 18F⁃FDG PET/CT and PET/MRI in pancreatic ductal adenocarcino⁃ ma[J].Abdom Radiol N Y,2018,43(2):415-434
参考文献 11
ZHANG G M,SUN H,SHI B,et al.Quantitative CT tex⁃ ture analysis for evaluating histologic grade of urothelial carcinoma[J].Abdom Radiol(NY),2017,42(2):561-568
参考文献 12
徐青青,单文莉,朱艳,等.基于CT影像组学对孤立性肺结节性质分类的预测效能[J].南京医科大学学报(自然科学版),2021,41(4):617-623
参考文献 13
EILAGHI A,BAIG S,ZHANG Y,et al.CT texture fea⁃ tures are associated with overall survival in pancreatic ductal adenocarcinoma ⁃ a quantitative analysis[J].BMC Med Imaging,2017,17(1):38
参考文献 14
邹金钊,叶靖,徐圆.CT纹理特征结合机器学习对膀胱尿路上皮癌T分期及病理级别的诊断价值[J].临床放射学杂志,2020,39(8):1553-1558
参考文献 15
WANG Y W,ZHANG X H,WANG B T,et al.Value of texture analysis of intravoxel incoherent motion parame⁃ ters in differential diagnosis of pancreatic neuroendocrine tumor and pancreatic adenocarcinoma[J].Chin Med Sci J,2019,34(1):1-9
参考文献 16
NAGTEGAAL I D,ODZE R D,KLIMSTRA D,et al.The 2019 WHO classification of tumours of the digestive sys⁃ tem[J].Histopathology,2020,76(2):182-188
参考文献 17
AL ⁃ HAWARY M M,FRANCIS I R,CHARI S T,et al.Pancreatic ductal adenocarcinoma radiology reporting template:consensus statement of the society of abdominal radiology and the American pancreatic association 1[J].Gastroenterology,2014,146(1):291-304
参考文献 18
PEREIRA J A,ROSADO E,BALI M,et al.Pancreatic neuroendocrine tumors:correlation between histogram analysis of apparent diffusion coefficient maps and tumor grade[J].Abdom Imaging,2015,40(8):3122-3128
参考文献 19
王利茹,侯兴宁,张建荣.肿瘤标记物 CA199、CA125、 CEA和D⁃二聚体对PDAC的早期诊断意义[J].宁夏医学杂志,2016,38(12):1190-1192
参考文献 20
姚浩琛.胰头癌胰十二指肠切除术后复发危险因素分析[D].长春:吉林大学,2018
参考文献 21
王中秋,黎介寿,卢光明,等.PDAC的CT增强和瘤体微血管密度及病理分级的相关性研究[J].中华医学杂志,2003,83(21):1882-1886
参考文献 22
REINERT C P,BAUMGARTNER K,HEPP T,et al.Com⁃ plementary role of computed tomography texture analysis for differentiation of pancreatic ductal adenocarcinoma from pancreatic neuroendocrine tumors in the portal ⁃ ve⁃ nous enhancement phase[J].Abdom Radiol(NY),2020,45(3):750-758
参考文献 23
孔令重,姜壮国,曹小江,等.Logistic 回归分析高分别率CT的纹理特征对孤立性肺结节诊断价值[J].新疆医科大学学报,2019,42(2):198-202
参考文献 24
沈力,征锦,叶靖,等.基于CT增强扫描图像的灰度直方图纹理分析评估PDAC分化程度[J].中华消化病与影像杂志(电子版),2019,9(1):8-13
目录contents

    摘要

    目的:探究基于术前增强纹理特征构建模型对评估胰腺导管腺癌(pancreatic ductal adenocarcinoma,PDAC)分化程度的价值。方法:回顾性收集2017年1月—2020年10月66例PDAC患者的病例资料,另外34例来自其他医院的PDAC患者被用于外部验证,根据术后病理结果分为高分化、中-低分化两组,分别记录患者的性别、年龄、肿瘤部位、肿瘤最大径、肿瘤强化程度、血管侵犯情况等临床及常规影像特征,进行单因素回归分析。采用ITK-SNAP软件勾画CT检查动、静脉期图像的感兴趣区(ROI),并提取图像纹理特征。利用单因素分析和二元 Logistic回归筛选独立预测因子并构建CT纹理特征模型,将训练组建立的预测模型直接应用于外部验证组,检验模型的准确度。应用受试者工作特征曲线(ROC)的曲线下面积(AUC)评价预测模型诊断价值。结果:基于动脉期及静脉期分别筛选出1个和2个纹理特征,分别为运行熵(run entropy)、区域百分比(zone percentage)和区域大小不均匀性(size-zone non-uniformity),其成为具有特征性的预测参数并分别构建了预测模型,基于CT动脉期纹理特征模型在训练组和验证组的AUC、灵敏度及特异度分别为0.716、0.581、0.824和0.722、0.600、0.765;基于CT静脉期纹理特征模型在训练组和验证组的AUC、灵敏度及特异度分别为0.895、0.781、0.882和0.873、0.722、0.929。结论:CT增强图像纹理特征在高分化、中-低分化PDAC之间存在差异,给术前评估PDAC恶性程度提供了新的方法。

    Abstract

    Objective:To investigate the value of the texture analysis based on preoperative enhanced CT in predicting the histopathological differentiation of pancreatic ductal adenocarcinoma(PDAC). Methods:Pathological data of 66 patients with PDAC from January 2017 to October 2020 were retrospectively collected,other 34 PDAC patients from other hospital were used for external validation cohort,and they were divided into two groups according to postoperative pathological results:high differentiation and moderately - poorly differentiation. The clinical and conventional imaging characteristics such as gender,age,tumor site,maximum tumor diameter,tumor enhancement and vascular invasion were recorded for univariate regression analysis. ROI based on both arterial phase and venous phase of the preoperative enhanced CT was automatically drawn by ITK -SNAP software and texture features were extracted. Univariate analysis was used to compare the texture features between the two groups,and the texture features with statistical difference were included in binary logistic regression model,and the prediction model of arterial phase and venous phase were established respectively. The prediction model established by the training group is directly applied to the external validation group. The AUC values of ROC were used to evaluate the diagnostic value of prediction model. Results:One and two texture features were selected to construct prediction model respectively based on the CT arterial and venous phase,including run entropy,zone percentage and size-zone non-uniformity. The AUC,sensitivity and specificity were 0.716,0.581,0.824,0.722,0.600,and 0.765 in the training group and the validation group based on the CT arterial texture feature model respectively. The AUC,sensitivity and specificity of the CT venous texture feature model in the training group and the validation group were 0.895,0.781,0.882,0.873,0.722,0.929, respectively. Conclusion:There are differences in the characteristic parameters of texture analysis of CT enhanced images between highly and moderately-poorly differentiated pancreatic ductal adenocarcinoma,providing a new method for preoperative evaluation of the malignant degree of pancreatic ductal adenocarcinoma carcinoma.

  • 胰腺导管腺癌(pancreatic ductal adenocarcinoma, PDAC)是临床常见的消化道肿瘤之一,PDAC早期诊断较为困难,大多数患者确诊时已处于中晚期且预后较差,5年生存率不足5%,90%的患者在确诊后1年内死亡[1-2]。尽管治疗方法有所进展,但PDAC的5年生存率并没有明显改善。由于PDAC预后较差,我们需要通过进一步的风险分层来制定更有效的治疗策略[3];研究表明PDAC的分化程度是PDAC术后生存的独立预后因素[4-5];与高分化PDAC相比,低分化PDAC具有更强的侵袭能力和更短的生存时间[6];而对于生存期较短的患者来说,手术切除相关风险远远大于益处[7]。Nurmi等[8] 发现,对于低分化PDAC,新辅助治疗可以提供比前期手术更长的生存时间。因此,如果能在术前有效评估PDAC的分化程度,对于患者治疗方案的选择和预后评估有着重要的意义。目前确定PDAC分化程度的方法主要包括超声内镜引导下的细针活检(endoscopic ultrasound ⁃ guided fine needle aspiration,EUS⁃FNA),但由于胰腺肿瘤的异质性,穿刺活检获得的标本体积有限,穿刺并不能获得肿瘤的分化程度[9]

  • 由于CT具有较高的时间和空间分辨率,是评估胰腺疾病的主要手段[10],但是CT检查评估结果受限于诊断医师的经验和专业能力,具有一定的主观性,并且获得的信息有限。基于CT图像的纹理分析是一种可以定量提取纹理信息的图像处理技术,可以用数学方法检测视觉上无法感知的像素强度变化,已经成为评估肿瘤异质性、预测治疗反应和预后的一种有前景的技术[11-12]。纹理分析已经被用于预测PDAC的预后以及PDAC与胰腺神经内分泌肿瘤的鉴别诊断等方面[13-15]。然而,目前应用纹理分析对PDAC病理分化程度预测的相关研究较少。本研究旨在探索基于术前增强CT图像的纹理特征预测胰PDAC理分化程度的作用。

  • 1 对象和方法

  • 1.1 对象

  • 回顾性分析2017年1月—2020年10月66例经病理证实并且有明确肿瘤分化程度评估的PDAC患者,纳入训练组,其中男36例,女30例,年龄范围31~87岁,平均(63.4 ±11.2)岁;并从南京市中医院选取了2017年10月—2021年8月34例PDAC患者作为外部验证组,其中男22例,女12例,年龄27~83岁,平均(58.1±10.9)岁。临床表现主要为上腹部不适、消化不良、黄疸。排除标准如下:①术前接受抗肿瘤治疗;②图像质量差,包括肿瘤内的伪影;③病灶体积太小,CT图像上显示不清;④没有完整的术前影像资料。

  • 1.2 方法

  • 1.2.1 病理分析

  • 所有手术切除的肿瘤均经福尔马林固定,石蜡包埋,苏木精⁃伊红染色。根据WHO消化系统肿瘤分类(第5版)[16] 分为:高分化腺癌、中分化腺癌、低分化腺癌,高分化腺癌由大量管样结构加上中等大小的肿瘤性腺体构成,中分化腺癌以埋于纤维间质中的中等大小、形状各异的导管样结构为主,低分化腺癌由密集排列的、形状不规则的小腺体以及完全取代腺泡组织的实性癌细胞巢或是条索结构构成。本研究中,由于病灶生长的不均质性,不同分化程度的成分可以同时存在同一病灶中,但在高分化肿瘤中出现低分化灶是很少见的,并且分化程度较低的成分更容易发生血管侵犯,预后较高分化病灶差,因此本研究的分化程度分组以整个病灶分化最低的部分为准[17]。根据术后病理分化程度,将本研究纳入的训练组病例分为两组:高分化组34例,中⁃低分化组32例,将纳入的外部验证组病例分为两组:高分化组21例,中⁃低分化组13例。本研究获得南京大学医学院附属鼓楼医院医学伦理委员会批准(2021⁃364⁃01)。

  • 1.2.2 CT扫描参数

  • 本研究66例患者均采用GE Light⁃speed 64排螺旋CT扫描仪,扫描参数如下:管电压120kVp,管电流200~250mAs,扫描层厚5mm,层间距5mm。所有患者在进行CT检查前至少禁食8h,均采用仰卧位进行扫描,范围:上至膈肌顶部,下至双肾下极。行CT平扫扫描后,以3.5mL/s的注射速率,由高压注射器经肘正中静脉注射对比剂(欧乃派克) 1.5mL/kg,注射后约30s和60s行动脉期和静脉期扫描。

  • 1.2.3 CT图像分析

  • 训练组及外部验证组CT图像通过影像存档和通信系统(PACS)传输到医生工作站后,分别由两位经验丰富的放射科医生(均具有10年以上经验)对以下影像学特征进行评估:①肿瘤的大小(肿瘤的最大层面测量肿瘤的直径,cm);②位置(头颈、钩突、体尾);③胰尾萎缩(胰尾体积明显缩小);④胰管扩张(主胰管直径≥3mm);⑤边界:边界清楚(肿瘤边缘清楚无毛刺或浸润少于80%)或边界模糊 (肿瘤边缘毛糙)[18];⑥胆总管扩张(胆总管最宽处直径≥9mm);⑦血管侵犯:观察肝总动脉、脾动静脉、胃十二指肠动脉、肠系膜上动静脉和门静脉受侵犯情况,标准指肿瘤与血管接触范围超过180°,可伴有血管的闭塞或血栓形成;⑧测量感兴趣区内肿瘤及胰腺实质的CT值,共测量3次,取其平均值,尽量选择动脉期及静脉期图像的同一层面进行测量。增强扫描肿瘤的强化程度使用肿瘤实质与正常胰腺组织的静脉期CT值的比值(tumor to pancreas ratio at portal venous phase,TPER)表示。两位医生对术后的病理结果及临床信息并不知情,当评估结果不一致时,以两位医生讨论后的结果为准。

  • 1.2.4 CT纹理分析

  • 从PACS获得所有患者的腹部增强CT动、静脉期图像,并以DICOM格式导出,本研究采用计算机自动分割算法对肿瘤感兴趣区域(region of interest, ROI)进行全容积分割,然后由上述两位放射科医师使用ITK⁃SNAP(开源软件,https://www.itksnap.org/) 对自动分割结果进行校正。避开肿瘤周围的胰腺实质组织、十二指肠、胰胆管及胰周血管,最终多层面ROI重建获得VOI(volume of interest)(图1)。将分割完毕的图像导入Pyradiomics软件(开源软件, https://github.com/Radiomics/pyradiomics)并从中提取影像组学特征732个,采用其中72个纹理特征,包括18个直方图特征、22个灰度共生矩阵特征(gray⁃level cooccurrence matrix,GLCM)、16个灰度游程长度矩阵特征(gray⁃level run length,GLRL) 及16个灰度区域大小矩阵特征(gray ⁃level zone size matrix,GLZSM)。

  • 1.3 统计学方法

  • 统计分析采用SPSS 22.0,所有计量资料使用正态性分布检验,符合正态分布的计量资料用均数±标准差(x-±s)表示,两组间比较采用两独立样本 t检验;不符合正态分布的计量资料用中位数(四分位数)[MP25P75)]表示,两组间比较采用Mann ⁃ Whitney U检验。计数资料以绝对数或百分比表示,组间比较采用卡方检验或连续校正卡方检验,将单因素分析有统计学差异的参数纳入多因素Logistic回归分析,进一步筛选出最佳预测参数并建立预测模型。通过受试者工作特征(receiver operating characteristic,ROC)曲线,得到模型及各参数曲线下面积(area under curve,AUC)、最佳临界值、灵敏度及特异度。以双侧P< 0.05为差异有统计学意义。

  • 2 结果

  • 2.1 临床和影像学特征

  • 本研究中训练组共66例PDAC患者,其中高分化34例,中⁃低分化32例。外部验证组共有34例PDAC患者,其中高分化21例,中低分化13例。训练组中⁃低分化PDAC患者血清糖类抗原125(carbohy⁃ drate antigen 125,CA125)水平显著高于高分化患者 (P=0.001),中⁃低分化PDAC在静脉期CT图像上肿瘤实质与正常胰腺组织静脉期CT值的比值(tumor to pancreas ratio at portal venous phase,TPER)明显低于高分化PDAC(P=0.005,表1)。验证组临床及影像特征在高、中低分化组间差异均无统计学意义 (P> 0.05,表2)。

  • 2.2 CT动、静脉期纹理特征的筛选和模型预测效能分析

  • 将选取的纹理特征进行单因素分析,基于动脉期、静脉期分别筛选出5、16个纹理特征。将上述CT纹理特征纳入多因素Logistic回归分析,并分别构建动、静脉期CT纹理特征模型,结果显示动脉期纹理特征中的运行熵(run entropy)和静脉期纹理特征中的区域百分比(zone percentage)及区域大小不均匀性(size⁃zone non⁃uniformity)是预测PDAC分化程度的最佳影响因素(P 值分别为0.024、0.013和0.016)。在训练组和验证组进行ROC曲线分析,基于CT动脉期纹理特征模型在训练组和验证组的AUC、灵敏度及特异度分别为0.716、0.581、0.824和0.722、0.600、0.765;基于CT静脉期纹理特征模型在训练组和验证组的AUC、灵敏度及特异度分别为0.895、0.781、0.882和0.873、0.722、0.929(表3)。绘制训练组及验证组纹理特征模型预测PDAC病理分化程度的ROC曲线(图2)。

  • 图1 胰腺导管腺癌患者增强CT检查的病灶勾画及重建

  • Fig.1 Focal delineation and reconstruction of ductal adenocarcinoma of the pancreas on enhanced CT

  • 表1 训练组高分化和中⁃低分化PDAC患者临床资料与常规影像特征

  • Table1 Clinical data and conventional imaging characteristics of patients with highly differentiated or moderately and poorly differentiated pancreatic ductal adenocarcinoma in the training group

  • TPER:肿瘤实质与正常胰腺组织的静脉期CT值的比值(tumor to pancreas ratio at portal venous phase)。

  • 表2 验证组高分化和中⁃低分化PDAC患者临床资料与常规影像特征

  • Table2 Clinical data and conventional imaging characteristics of patients with highly differentiated or moderately and poorly differentiated pancreatic ductal adenocarcinoma in the training group

  • 3 讨论

  • 本研究建立了基于术前CT动、静脉纹理特征预测术后高分化PDAC及中⁃低分化PDAC模型,其中动脉期型的AUC值为0.716,灵敏度为0.581,特异度为0.824。静脉期模型的AUC值为0.895,灵敏度为0.781,特异度为0.882。结果表明,对于预期寿命较短的PDAC患者,利用CT纹理分析有助于术前风险分层,进而最大化患者生存期。

  • 血清中CA125最早应用于卵巢癌的诊断,但在消化系统肿瘤(PDAC、结直肠癌、胃癌)患者血清中也均有表达[19]。以往有研究表明,CA125对于发现早期PDAC有一定的意义,是一种较好的PDAC肿瘤标志物[20]。本研究发现中⁃低分化PDAC CA125水平高于高分化PDAC,差异有统计学意义(P< 0.05),这一发现可能为今后研究PDAC的分化程度提供新的理论依据。

  • 表3 CT动、静脉期纹理特征模型预测PDAC分化程度效能

  • Table3 Efficacy of CT arteriovenous texture model in predicting differentiation degree of pancreatic ductal adenocarcinoma

  • 图2 多参数模型预测中⁃低分化PDAC的ROC曲线图

  • Fig.2 ROC curve of multi⁃parameter model for prediction of medium⁃low differentiation pancreatic cancer

  • CT增强检查是PDAC术前检查最常用的影像学方法,王中秋等[21] 评估了32例PDAC患者的术前增强CT检查,研究结果表明PDAC的分化程度和病灶实质期的CT增强程度呈正比,即PDAC的分化程度越高,其病灶的CT增强程度就越高,PDAC的CT增强程度与病灶内不同成分的比例有关,而病灶的成分比例因肿瘤的分化程度高低而不同。本研究发现中⁃低分化PDAC的强化程度明显低于高分化PDAC,但是CT检查评估结果受限于诊断医师的经验和专业能力,具有一定的主观性,容易产生误差。研究发现PDAC纹理分析技术通过数学方法检测视觉上无法感知的像素强度变化,以反映病变内的异质性和病变之间的细小差异[11],已被证实在PDAC诊断和预后中具有较高的价值[22]

  • 本研究最终筛选出3个CT纹理特征,分别为运行熵、区域百分比和区域大小不均匀。运行熵测量运行长度的随机分布程度,其值越高,表示纹理的异质性越高;区域百分比是通过区域数量和体素数量的比例来量化纹理的粗糙度,该值越高,表明图像的纹理越精细;区域大小不均匀性测量图像中尺寸区体积的变异性,该值越低,说明图像中不同区域的纹理之间缺少变化、质地均匀。中⁃低分化PDAC在动脉期图像上运行熵值明显较高,同时在静脉期图像上区域百分比值明显较低,而在区域大小不均匀性的值明显较高,因此,中⁃低分化PDAC在CT图像上像素灰度变化幅度较大、分布规则较乱。孔令重等[23] 研究发现肿瘤内异质性的一个重要组成部分是肿瘤血管分布的不规则性。肿瘤血管的畸形、不规则结构导致肿瘤出现囊变、坏死。我们推测中⁃低分化PDAC的图像不均匀性与肿瘤内部的囊变坏死有关,也表明中⁃低分化PDAC异质性较高分化PDAC高,从而也预示着较差的预后。

  • 本研究基于CT动、静脉期增强图像建立PDAC病理学分化程度预测模型,结果显示静脉期较动脉期预测效能稍高,可能是因为胰腺实质在静脉期与周围正常组织有明显的对比,有助于更准确地勾画ROI。沈力等[24] 回顾性分析了49例PDAC病例,提取了静脉期灰度直方图特征,发现纹理特征中均值的AUC值最大(AUC=0.695),具有最佳的诊断效能,灵敏度和特异度分别为0.864和0.444。本研究建立的预测模型诊断效能又有了很大的提高,可能是因为本研究在灰度直方图纹理特征的基础上加入了灰度共生(gray⁃level cooccurrence matrix, GLCM)、灰度游程长度(gray⁃level run length,GLRL) 和灰度区域大小(gray⁃level zone size matrix,GLZSM) 纹理特征,同时本研究分别基于动脉期、静脉期图像进行了ROI的勾画,而且测量了整个肿瘤的体积,这样可以减少选择单一层面带来的误差,能够更加全面地反映肿瘤的异质性。此外,本研究在图像的勾画上也有明显的优势,肿瘤的ROI勾画采用了一种新的深度学习网络系统自动分割方法,这种新的分割方法有助于得到较清晰的分割边界,从而显著提高了影像学特性的稳定性。尽管存在这些差异,但两项研究都支持一个相似的结论,即CT纹理分析是预测PDAC分化程度的一种有前景且无创的方法。

  • 本研究尚存在一些局限性:①纹理分析需要大量的样本,本研究纳入的病例数仍然有限,未来需要收集更多的病例来提高预测效能;②本研究为回顾性设计,可能会引入选择偏差;③只考察了纹理学特征用于预测PDAC的病理学分化程度。在未来的研究中,我们应该结合基因组学特征和蛋白质组学等特征,其预测PDAC分化程度的诊断效能将进一步提高。

  • 综上所述,利用术前CT动静脉期增强图像纹理分析的特征参数建立的预测模型在高、中⁃低分化PDAC之间存在差异,给术前评估PDAC的恶性程度提供了新的方法。

  • 参考文献

    • [1] KARDOSH A,LICHTENSZTAJN D,GUBENS M,et al.Long ⁃ term survivors of pancreatic cancer:a California population⁃based study[J].Pancreas,2018,47(8):958-966

    • [2] NEESSE A,BAUER C A,ÖHLUND D,et al.Stromal biol⁃ ogy and therapy in pancreatic cancer:ready for clinical translation?[J].Gut,2019,68(1):159-171

    • [3] WASIF N,KO C Y,FARRELL J,et al.Impact of tumor grade on prognosis in pancreatic cancer:should we in⁃ clude grade in AJCC staging?[J].Ann Surg Oncol,2010,17(9):2312-2320

    • [4] JEEKEL H.Prognostic factors following curative resec⁃ tion for pancreatic adenocarcinoma[J].Ann Surg,2004,240(2):384

    • [5] WINTER J,CAMERON J,CAMPBELL K.1423 pancreat⁃ icoduodenectomies for pancreatic cancer:a single⁃institu⁃ tion experience[J].J Gastrointest Surg,2006,10(9):1199-1121

    • [6] MACÍAS N,SAYAGUÉS J M,ESTEBAN C,et al.Histo⁃ logic tumor grade and preoperative bilary drainage are the unique independent prognostic factors of survival in pan⁃ creatic ductal adenocarcinoma patients after pancreatico⁃ duodenectomy[J].J Clin Gastroenterol,2018,52(2):e11-e17

    • [7] CASSINOTTO C,CHONG J,ZOGOPOULOS G,et al.Re⁃ sectable pancreatic adenocarcinoma:role of CT quantita⁃ tive imaging biomarkers for predicting pathology and pa⁃ tient outcomes[J].Eur J Radiol,2017,90:152-158

    • [8] NURMI A,MUSTONEN H,PARVIAINEN H,et al.Neo⁃ adjuvant therapy offers longer survival than upfront sur⁃ gery for poorly differentiated and higher stage pancreatic cancer[J].Acta Oncol,2018,57(6):799-806

    • [9] GERLINGER M,ROWAN A J,HORSWELL S,et al.In⁃ tratumor heterogeneity and branched evolution revealed by multiregion sequencing[J].N Engl J Med,2012,366(10):883-892

    • [10] YEH R,DERCLE L,GARG I,et al.The role of 18F⁃FDG PET/CT and PET/MRI in pancreatic ductal adenocarcino⁃ ma[J].Abdom Radiol N Y,2018,43(2):415-434

    • [11] ZHANG G M,SUN H,SHI B,et al.Quantitative CT tex⁃ ture analysis for evaluating histologic grade of urothelial carcinoma[J].Abdom Radiol(NY),2017,42(2):561-568

    • [12] 徐青青,单文莉,朱艳,等.基于CT影像组学对孤立性肺结节性质分类的预测效能[J].南京医科大学学报(自然科学版),2021,41(4):617-623

    • [13] EILAGHI A,BAIG S,ZHANG Y,et al.CT texture fea⁃ tures are associated with overall survival in pancreatic ductal adenocarcinoma ⁃ a quantitative analysis[J].BMC Med Imaging,2017,17(1):38

    • [14] 邹金钊,叶靖,徐圆.CT纹理特征结合机器学习对膀胱尿路上皮癌T分期及病理级别的诊断价值[J].临床放射学杂志,2020,39(8):1553-1558

    • [15] WANG Y W,ZHANG X H,WANG B T,et al.Value of texture analysis of intravoxel incoherent motion parame⁃ ters in differential diagnosis of pancreatic neuroendocrine tumor and pancreatic adenocarcinoma[J].Chin Med Sci J,2019,34(1):1-9

    • [16] NAGTEGAAL I D,ODZE R D,KLIMSTRA D,et al.The 2019 WHO classification of tumours of the digestive sys⁃ tem[J].Histopathology,2020,76(2):182-188

    • [17] AL ⁃ HAWARY M M,FRANCIS I R,CHARI S T,et al.Pancreatic ductal adenocarcinoma radiology reporting template:consensus statement of the society of abdominal radiology and the American pancreatic association 1[J].Gastroenterology,2014,146(1):291-304

    • [18] PEREIRA J A,ROSADO E,BALI M,et al.Pancreatic neuroendocrine tumors:correlation between histogram analysis of apparent diffusion coefficient maps and tumor grade[J].Abdom Imaging,2015,40(8):3122-3128

    • [19] 王利茹,侯兴宁,张建荣.肿瘤标记物 CA199、CA125、 CEA和D⁃二聚体对PDAC的早期诊断意义[J].宁夏医学杂志,2016,38(12):1190-1192

    • [20] 姚浩琛.胰头癌胰十二指肠切除术后复发危险因素分析[D].长春:吉林大学,2018

    • [21] 王中秋,黎介寿,卢光明,等.PDAC的CT增强和瘤体微血管密度及病理分级的相关性研究[J].中华医学杂志,2003,83(21):1882-1886

    • [22] REINERT C P,BAUMGARTNER K,HEPP T,et al.Com⁃ plementary role of computed tomography texture analysis for differentiation of pancreatic ductal adenocarcinoma from pancreatic neuroendocrine tumors in the portal ⁃ ve⁃ nous enhancement phase[J].Abdom Radiol(NY),2020,45(3):750-758

    • [23] 孔令重,姜壮国,曹小江,等.Logistic 回归分析高分别率CT的纹理特征对孤立性肺结节诊断价值[J].新疆医科大学学报,2019,42(2):198-202

    • [24] 沈力,征锦,叶靖,等.基于CT增强扫描图像的灰度直方图纹理分析评估PDAC分化程度[J].中华消化病与影像杂志(电子版),2019,9(1):8-13

  • 参考文献

    • [1] KARDOSH A,LICHTENSZTAJN D,GUBENS M,et al.Long ⁃ term survivors of pancreatic cancer:a California population⁃based study[J].Pancreas,2018,47(8):958-966

    • [2] NEESSE A,BAUER C A,ÖHLUND D,et al.Stromal biol⁃ ogy and therapy in pancreatic cancer:ready for clinical translation?[J].Gut,2019,68(1):159-171

    • [3] WASIF N,KO C Y,FARRELL J,et al.Impact of tumor grade on prognosis in pancreatic cancer:should we in⁃ clude grade in AJCC staging?[J].Ann Surg Oncol,2010,17(9):2312-2320

    • [4] JEEKEL H.Prognostic factors following curative resec⁃ tion for pancreatic adenocarcinoma[J].Ann Surg,2004,240(2):384

    • [5] WINTER J,CAMERON J,CAMPBELL K.1423 pancreat⁃ icoduodenectomies for pancreatic cancer:a single⁃institu⁃ tion experience[J].J Gastrointest Surg,2006,10(9):1199-1121

    • [6] MACÍAS N,SAYAGUÉS J M,ESTEBAN C,et al.Histo⁃ logic tumor grade and preoperative bilary drainage are the unique independent prognostic factors of survival in pan⁃ creatic ductal adenocarcinoma patients after pancreatico⁃ duodenectomy[J].J Clin Gastroenterol,2018,52(2):e11-e17

    • [7] CASSINOTTO C,CHONG J,ZOGOPOULOS G,et al.Re⁃ sectable pancreatic adenocarcinoma:role of CT quantita⁃ tive imaging biomarkers for predicting pathology and pa⁃ tient outcomes[J].Eur J Radiol,2017,90:152-158

    • [8] NURMI A,MUSTONEN H,PARVIAINEN H,et al.Neo⁃ adjuvant therapy offers longer survival than upfront sur⁃ gery for poorly differentiated and higher stage pancreatic cancer[J].Acta Oncol,2018,57(6):799-806

    • [9] GERLINGER M,ROWAN A J,HORSWELL S,et al.In⁃ tratumor heterogeneity and branched evolution revealed by multiregion sequencing[J].N Engl J Med,2012,366(10):883-892

    • [10] YEH R,DERCLE L,GARG I,et al.The role of 18F⁃FDG PET/CT and PET/MRI in pancreatic ductal adenocarcino⁃ ma[J].Abdom Radiol N Y,2018,43(2):415-434

    • [11] ZHANG G M,SUN H,SHI B,et al.Quantitative CT tex⁃ ture analysis for evaluating histologic grade of urothelial carcinoma[J].Abdom Radiol(NY),2017,42(2):561-568

    • [12] 徐青青,单文莉,朱艳,等.基于CT影像组学对孤立性肺结节性质分类的预测效能[J].南京医科大学学报(自然科学版),2021,41(4):617-623

    • [13] EILAGHI A,BAIG S,ZHANG Y,et al.CT texture fea⁃ tures are associated with overall survival in pancreatic ductal adenocarcinoma ⁃ a quantitative analysis[J].BMC Med Imaging,2017,17(1):38

    • [14] 邹金钊,叶靖,徐圆.CT纹理特征结合机器学习对膀胱尿路上皮癌T分期及病理级别的诊断价值[J].临床放射学杂志,2020,39(8):1553-1558

    • [15] WANG Y W,ZHANG X H,WANG B T,et al.Value of texture analysis of intravoxel incoherent motion parame⁃ ters in differential diagnosis of pancreatic neuroendocrine tumor and pancreatic adenocarcinoma[J].Chin Med Sci J,2019,34(1):1-9

    • [16] NAGTEGAAL I D,ODZE R D,KLIMSTRA D,et al.The 2019 WHO classification of tumours of the digestive sys⁃ tem[J].Histopathology,2020,76(2):182-188

    • [17] AL ⁃ HAWARY M M,FRANCIS I R,CHARI S T,et al.Pancreatic ductal adenocarcinoma radiology reporting template:consensus statement of the society of abdominal radiology and the American pancreatic association 1[J].Gastroenterology,2014,146(1):291-304

    • [18] PEREIRA J A,ROSADO E,BALI M,et al.Pancreatic neuroendocrine tumors:correlation between histogram analysis of apparent diffusion coefficient maps and tumor grade[J].Abdom Imaging,2015,40(8):3122-3128

    • [19] 王利茹,侯兴宁,张建荣.肿瘤标记物 CA199、CA125、 CEA和D⁃二聚体对PDAC的早期诊断意义[J].宁夏医学杂志,2016,38(12):1190-1192

    • [20] 姚浩琛.胰头癌胰十二指肠切除术后复发危险因素分析[D].长春:吉林大学,2018

    • [21] 王中秋,黎介寿,卢光明,等.PDAC的CT增强和瘤体微血管密度及病理分级的相关性研究[J].中华医学杂志,2003,83(21):1882-1886

    • [22] REINERT C P,BAUMGARTNER K,HEPP T,et al.Com⁃ plementary role of computed tomography texture analysis for differentiation of pancreatic ductal adenocarcinoma from pancreatic neuroendocrine tumors in the portal ⁃ ve⁃ nous enhancement phase[J].Abdom Radiol(NY),2020,45(3):750-758

    • [23] 孔令重,姜壮国,曹小江,等.Logistic 回归分析高分别率CT的纹理特征对孤立性肺结节诊断价值[J].新疆医科大学学报,2019,42(2):198-202

    • [24] 沈力,征锦,叶靖,等.基于CT增强扫描图像的灰度直方图纹理分析评估PDAC分化程度[J].中华消化病与影像杂志(电子版),2019,9(1):8-13