-
复发性急性胰腺炎(recurrent acute pancreatitis,RAP)是外科常见的急腹症之一,具有发病急、进展快、病死率高等特点,发病的原因包括胆结石或胆汁淤积、饮酒、Oddi括约肌功能障碍、基因突变、胰管干扰胰腺分泌物的正常流动等[1-4]。诊断为急性胰腺炎的患者的复发率为17%~22%,高达36%的急性胰腺炎患者最终会发展为慢性胰腺炎(chronic pancreatitis,CP)[5-6]。影像检查在急性胰腺炎的诊断中具有重要的作用,其中超导MRI与多层螺旋CT最为常用,而CT具有图像采集速度快、空间分辨率高、可重复性强、操作简便等优势,更具推广价值。但RAP患者在急性胰腺炎发作消退后及复发发作期间,其胰腺在CT上通常显示正常,这使得RAP的诊断具有挑战性[7-9]。因此,需要诊断工具来进一步完善RAP的临床诊断,并将其与其他腹痛区分开来。
放射组学是应用计算机图像处理的手段将ROI的影像数据转化为可挖掘的高维特征数据,利用纹理分析对放射成像的定量特征进行研究,可提供许多疾病过程中潜在的病理学外的信息,其已经在乳腺癌、肺癌和脑转移等疾病中显示出潜在的价值[10-11]。由于腹痛和胰酶升高都不是急性胰腺炎的专属特征,因此,从CT图像中提取出的放射组学特征在诊断胰腺炎患者的病变特征方面有重要意义。本研究通过观察RAP患者的胰腺CT放射组学特征,分析胰腺的放射组学特征是否能区分功能性腹痛(functional abdominal pain,FAP)、RAP和CP患者,以期为CT成像的放射组学特征在胰腺炎精准诊断中的应用提供理论依据。
-
与RAP和FAP患者相比,CP患者平均年龄大9岁左右,糖尿病发病率较高,且重度饮酒者和当前吸烟者占比较多(表1)。
基线特征 功能性腹痛
(n=16)复发性急性胰
腺炎(n=18)慢性胰腺炎
(n=14)年龄( ,岁)$\bar x\pm s $ 45.4±10.1 43.8±12.3 54.3±11.5 男性[例(%)] 5(31.3) 10(55.6) 8(57.1) 体重指数( ,kg/m2)$\bar x\pm s $ 25.4±3.6 23.5±4.3 22.7±3.8 糖尿病[例(%)] 2(12.5) 3(16.7) 6(42.8) 饮酒程度[例(%)] 重度 3(18.8) 4(22.2) 7(50.0) 中度 1(6.3) 2(11.1) 1(7.1) 轻微 5(31.3) 2(11.1) 1(7.1) 节制 7(43.8) 10(55.6) 5(35.7) 吸烟史[例(%)] 当前吸烟 2(12.5) 3(16.7) 3(21.4) 过去吸烟 1(6.3) 5(27.8) 3(21.4) 不吸烟 13(81.3) 10(55.6) 8(57.1) 表 1 临床确诊的不同类型腹痛患者的基线特征
Table 1. Baseline characteristics of patients with different types of abdominal pain based on clinical diagnosis
-
在单变量分析中,有9个放射组学特征在3组患者中的差异有统计学意义(Z=3.45~29.76,均P<0.05)(表2),其中8个是GLCM特征,1个是NGTDM特征。
放射组学特征 功能性腹痛(n=16) 复发性急性胰腺炎(n=18) 慢性胰腺炎(n=14) GLCM 集群突出 426.76±212.31 60.89±103.07 (22.3±4.26)×103 集群阴影 −0.46±1.24 −3.26±3.21 513.37±634.28 集群趋势 1.16±0.63 1.83±0.78 27.50±11.23 相关性 0.24±0.08 0.49±0.07 0.63±0.11 熵值 0.68±0.34 1.46±0.37 2.26±0.48 能量 0.38±0.07 0.23±0.06 0.17±0.03 同质性 207.23±86.37 0.78±0.05 0.63±0.04 相关性的
信息量度0.76±0.11 0.49±0.08 0.62±0.10 NGTDM 对比度 (0.52±0.98)×10−5 (2.06±1.03)×10−5 (4.23±2.11)×10−5 注:GLCM为灰度共生矩阵;NGTDM为邻域灰度差矩阵 表 2 临床确诊的不同类型腹痛患者的重要放射组学特征(
)$ \bar x \pm s $ Table 2. Important radiomics features of patients with different types of abdominal pain based on clinical diagnosis (
)$ \bar x \pm s $ -
对于这9个放射特征,我们构建了单独的ROC曲线来测量AUC;将RAP患者与FAP和CP患者分别进行比较时,AUC的范围分别为0.76~0.93和0.73~0.91(表3)。
放射组学特征 复发性急性胰腺炎对
功能性腹痛复发性急性胰腺炎对
慢性胰腺炎AUC P值 AUC P值 GLCM 集群突出 0.87 <0.001 0.87 <0.001 集群阴影 0.85 <0.001 0.91 <0.001 集群趋势 0.76 0.005 0.87 <0.001 相关性 0.91 <0.001 0.79 0.007 熵值 0.88 <0.001 0.81 0.006 能量 0.79 0.004 0.73 0.013 同质性 0.85 <0.001 0.75 0.008 相关性的信息量度 0.93 <0.001 0.81 0.006 NGTDM 对比度 0.85 <0.001 0.75 0.008 注:AUC为曲线下面积;GLCM为灰度共生矩阵;NGTDM为邻域灰度差矩阵;P值基于Wilcoxon秩和检验 表 3 不同类型腹痛患者组间比较中单个放射组学特征的曲 线下面积和P值
Table 3. Areas under the curve and P values of individual radiation features in patients with different types of abdominal pain were compared
IsoSVM机器学习模型的总体预测准确率为82.1%。FAP组的灵敏度、特异度分别为78.7%、100%,AUC为0.90。RAP组的灵敏度、特异度分别为95.2%、77.8%,AUC为 0.87,而CP组的灵敏度、特异度分别为70.9%、94.8%,AUC 为0.89(图1)。
CT成像的放射组学特征在胰腺炎诊断中的评估效能
Evaluation efficacy of radiomic features of CT imaging in the diagnosis of pancreatitis
-
摘要:
目的 探讨胰腺CT成像的放射组学特征在功能性腹痛(FAP)、复发性急性胰腺炎(RAP)、慢性胰腺炎(CP)患者诊断中的评估效能。 方法 回顾性分析2017年12月至2020年12月在首都医科大学附属北京天坛医院接受腹部增强CT检查的168例患者的CT影像资料,根据纳排标准,共选取48例患者进行研究,其中男性 23例、女性25例,年龄39~84(47.8±10.2)岁;FAP患者16例(FAP组)、RAP患者18例(RAP组)、CP患者14例(CP组)。通过在CT图像上绘制感兴趣区来勾画胰腺轮廓。从每个感兴趣区提取62个放射组学特征,共分为5类,分别是一阶统计量、灰度共生矩阵(GLCM)、灰度行程矩阵(GLRLM)、邻域灰度差矩阵(NGTDM)和形态学特征,并在3组患者间进行比较。根据组别建立3个IsoSVM机器学习模型,对IsoSVM模型进行训练,并在遗漏的交叉验证样本上进行测试。RAP患者、FAP患者、CP患者的放射组学特征的比较采用Wilcoxon秩和检验。使用受试者工作特征曲线及曲线下面积(AUC)评估个体放射组学特征的评估效能。 结果 在单变量分析中,9个放射组学特征(8个GLCM特征和1个NGTDM特征)在患者组间的差异有统计学意义(Z=3.45~29.76,均P<0.05)。RAP患者与FAP和CP患者分别进行比较,放射组学特征的AUC范围分别为0.76~0.93和0.73~0.91。IsoSVM机器学习模型的总体预测准确率为82.1%。FAP组的灵敏度、特异度分别为78.7%、100%,AUC为0.90。RAP组的灵敏度、特异度分别为95.2%、77.8%,AUC为 0.87,而CP组的灵敏度、特异度分别为70.9%、94.8%,AUC 为0.89。 结论 CT成像的部分放射组学特征对胰腺炎的诊断有较好的评估效能,可以区分FAP、RAP和CP患者。 -
关键词:
- 体层摄影术,X线计算机 /
- 放射组学 /
- 复发性急性胰腺炎 /
- 慢性胰腺炎
Abstract:Objective To evaluate the diagnostic efficacy of the radiological features of pancreatic CT imaging in patients with functional abdominal pain (FAP), recurrent acute pancreatitis (RAP), and chronic pancreatitis (CP). Methods Retrospective analysis was performed on the CT image data of 168 patients who received abdominal enhanced CT examination in Beijing Tiantan Hospital, Capital Medical University from December 2017 to December 2020. According to the criteria for admission and emission, 48 patients were selected for the study, including 23 males and 25 females, aged 39 to 84 (47.8±10.2) years, and 16 cases of FAP (FAP group), 18 cases of RAP (RAP group), and 14 cases of CP (CP group). The pancreas outline was obtained by drawing the region of interest on the CT image. Sixty-two radiologic features were extracted from each region of interest, which were divided into five categories, namely, the first-order statistics, the gray-level co-occurrence matrix (GLCM), the gray-level run-length matrix, the neighbouring gray tone difference matrix (NGTDM), and the morphological features, and compared among the three groups. According to the groups, three IsoSVM machine learning models were established, trained, and tested on the missing cross validation samples. The Wilcoxon rank sum test was used to compare the radiation characteristics of the patients with RAP, FAP, and CP. The predictive performance of individual radiological characteristics was evaluated using the receiver operator characteristic curve and the area under the curve (AUC). Results In the univariate analysis, a significant difference was found between the patient groups in nine radiation group characteristics (eight GLCM characteristics and one NGTDM characteristic) (Z=3.45–29.76, all P<0.05). Compared RAP patients with FAP and CP patients, the AUC ranges were 0.76–0.93 and 0.73–0.91. The overall prediction accuracy of the IsoSVM machine learning model was 82.1%. The sensitivity and specificity of the FAP group were 78.7% and 100%, respectively, and the AUC was 0.90. The sensitivity and specificity of the RAP group were 95.2% and 77.8%, respectively, and the AUC was 0.87, while those of the CP group were 70.9%, 94.8%, and 0.89, respectively. Conclusion Some of the radiographic features of CT imaging have a good evaluation efficiency in the diagnosis of pancreatitis and can distinguish between patients with FAP, RAP, and CP. -
表 1 临床确诊的不同类型腹痛患者的基线特征
Table 1. Baseline characteristics of patients with different types of abdominal pain based on clinical diagnosis
基线特征 功能性腹痛
(n=16)复发性急性胰
腺炎(n=18)慢性胰腺炎
(n=14)年龄( ,岁)$\bar x\pm s $ 45.4±10.1 43.8±12.3 54.3±11.5 男性[例(%)] 5(31.3) 10(55.6) 8(57.1) 体重指数( ,kg/m2)$\bar x\pm s $ 25.4±3.6 23.5±4.3 22.7±3.8 糖尿病[例(%)] 2(12.5) 3(16.7) 6(42.8) 饮酒程度[例(%)] 重度 3(18.8) 4(22.2) 7(50.0) 中度 1(6.3) 2(11.1) 1(7.1) 轻微 5(31.3) 2(11.1) 1(7.1) 节制 7(43.8) 10(55.6) 5(35.7) 吸烟史[例(%)] 当前吸烟 2(12.5) 3(16.7) 3(21.4) 过去吸烟 1(6.3) 5(27.8) 3(21.4) 不吸烟 13(81.3) 10(55.6) 8(57.1) 表 2 临床确诊的不同类型腹痛患者的重要放射组学特征(
)$ \bar x \pm s $ Table 2. Important radiomics features of patients with different types of abdominal pain based on clinical diagnosis (
)$ \bar x \pm s $ 放射组学特征 功能性腹痛(n=16) 复发性急性胰腺炎(n=18) 慢性胰腺炎(n=14) GLCM 集群突出 426.76±212.31 60.89±103.07 (22.3±4.26)×103 集群阴影 −0.46±1.24 −3.26±3.21 513.37±634.28 集群趋势 1.16±0.63 1.83±0.78 27.50±11.23 相关性 0.24±0.08 0.49±0.07 0.63±0.11 熵值 0.68±0.34 1.46±0.37 2.26±0.48 能量 0.38±0.07 0.23±0.06 0.17±0.03 同质性 207.23±86.37 0.78±0.05 0.63±0.04 相关性的
信息量度0.76±0.11 0.49±0.08 0.62±0.10 NGTDM 对比度 (0.52±0.98)×10−5 (2.06±1.03)×10−5 (4.23±2.11)×10−5 注:GLCM为灰度共生矩阵;NGTDM为邻域灰度差矩阵 表 3 不同类型腹痛患者组间比较中单个放射组学特征的曲 线下面积和P值
Table 3. Areas under the curve and P values of individual radiation features in patients with different types of abdominal pain were compared
放射组学特征 复发性急性胰腺炎对
功能性腹痛复发性急性胰腺炎对
慢性胰腺炎AUC P值 AUC P值 GLCM 集群突出 0.87 <0.001 0.87 <0.001 集群阴影 0.85 <0.001 0.91 <0.001 集群趋势 0.76 0.005 0.87 <0.001 相关性 0.91 <0.001 0.79 0.007 熵值 0.88 <0.001 0.81 0.006 能量 0.79 0.004 0.73 0.013 同质性 0.85 <0.001 0.75 0.008 相关性的信息量度 0.93 <0.001 0.81 0.006 NGTDM 对比度 0.85 <0.001 0.75 0.008 注:AUC为曲线下面积;GLCM为灰度共生矩阵;NGTDM为邻域灰度差矩阵;P值基于Wilcoxon秩和检验 -
[1] 闫媛媛, 张洁, 靳二虎. 改良CT和MR严重指数在急性胰腺炎诊断中的应用研究[J]. 临床和实验医学杂志, 2018, 17(10): 1037−1039. DOI: 10.3969/j.issn.1671-4695.2018.10.010.
Yan YY, Zhang J, Jin EH. Study on the application of modified CT and MR severity indices in patients with acute pancreatitis[J]. J Clin Exp Med, 2018, 17(10): 1037−1039. DOI: 10.3969/j.issn.1671-4695.2018.10.010.[2] 黄昊苏. 生物碱对急性胰腺炎的治疗作用及机制研究进展[J]. 中国普通外科杂志, 2018, 27(3): 367−373. DOI: 10.3978/j.issn.1005-6947.2018.03.016.
Huang HS. Research progress of therapeutic effect of alkaloids on acute pancreatitis and their mechanisms[J]. Chin J Gen Surg, 2018, 27(3): 367−373. DOI: 10.3978/j.issn.1005-6947.2018.03.016.[3] 黄莺, 李奇林, 张存程, 等. 老年急性胰腺炎彩色多普勒超声和多层螺旋CT检查的影像学表现及检出率比较[J]. 中国老年学杂志, 2018, 38(8): 1854−1856. DOI: 10.3969/j.issn.1005-9202.2018.08.027.
Huang Y, Li QL, Zhang CC, et al. Comparison of imaging findings and detection rate between color Doppler ultrasonography and multi-slice spiral CT in elderly patients with acute pancreatitis[J]. Chin J Gerontol, 2018, 38(8): 1854−1856. DOI: 10.3969/j.issn.1005-9202.2018.08.027.[4] Testoni PA. Acute recurrent pancreatitis: etiopathogenesis, diagnosis and treatment[J]. World J Gastroenterol, 2014, 20(45): 16891−16901. DOI: 10.3748/wjg.v20.i45.16891. [5] Sankaran SJ, Xiao AY, Wu LM, et al. Frequency of progression from acute to chronic pancreatitis and risk factors: a meta-analysis[J]. Gastroenterology, 2015, 149(6): 1490−1500.e1. DOI: 10.1053/j.gastro.2015.07.066. [6] Ahmed Ali U, Issa Y, Hagenaars JC, et al. Risk of recurrent pancreatitis and progression to chronic pancreatitis after a first episode of acute pancreatitis[J]. Clin Gastroenterol Hepatol, 2016, 14(5): 738−746. DOI: 10.1016/j.cgh.2015.12.040. [7] Jagannath S, Garg PK. Recurrent acute pancreatitis: current concepts in the diagnosis and management[J]. Curr Treat Options Gastroenterol, 2018, 16(4): 449−465. DOI: 10.1007/s11938-018-0196-9. [8] Coté GA, Imperiale TF, Schmidt SE, et al. Similar efficacies of biliary, with or without pancreatic, sphincterotomy in treatment of idiopathic recurrent acute pancreatitis[J]. Gastroenterology, 2012, 143(6): 1502−1509.e1. DOI: 10.1053/j.gastro.2012.09.006. [9] Wu BU, Johannes RS, Sun X, et al. The early prediction of mortality in acute pancreatitis: a large population-based study[J]. Gut, 2008, 57(12): 1698−1703. DOI: 10.1136/gut.2008.152702. [10] Parekh VS, Jacobs MA. Integrated radiomic framework for breast cancer and tumor biology using advanced machine learning and multiparametric MRI[J/OL]. NPJ Breast Cancer, 2017, 3: 43[2021-11-30]. https://www.nature.com/articles/s41523-017-0045-3. DOI: 10.1038/s41523-017-0045-3. [11] Peng LK, Parekh V, Huang P, et al. Distinguishing true progression from radionecrosis after stereotactic radiation therapy for brain metastases with machine learning and radiomics[J]. Int J Radiat Oncol Biol Phys, 2018, 102(4): 1236−1243. DOI: 10.1016/j.ijrobp.2018.05.041. [12] 中华医学会, 中华医学会杂志社, 中华医学会消化病学分会, 等. 急性胰腺炎基层诊疗指南(2019年)[J]. 中华全科医师杂志, 2019, 18(9): 819−826. DOI: 10.3760/cma.j.issn.1671-7368.2019.09.004.
Chinese Medical Association, Chinese Medical Journals Publishing House, Chinese Society of Gastroenterology, et al. Guideline for primary care of acute pancreatitis (2019)[J]. Chin J Gen Pract, 2019, 18(9): 819−826. DOI: 10.3760/cma.j.issn.1671-7368.2019.09.004.[13] 中国医师协会胰腺病专业委员会慢性胰腺炎专委会. 慢性胰腺炎诊治指南(2018年, 广州)[J]. 中华消化杂志, 2018, 38(11): 739−746. DOI: 10.3760/cma.j.issn.0254-1432.2018.11.003.
Special Committee of Chronic Pancreatitis of Pancreatic Disease Professional Committee of Chinese Medical Association. Guidelines for diagnosis and treatment of chronic pancreatitis (Guangzhou, 2018)[J]. Chin J Dig, 2018, 38(11): 739−746. DOI: 10.3760/cma.j.issn.0254-1432.2018.11.003.[14] 刘劲松. 功能性胃肠病罗马Ⅲ标准解读: 功能性腹痛[J]. 临床消化病杂志, 2006, 18(5): 266−267. DOI: 10.3870/j.issn.1005-541X.2006.05.007.
Liu JS. Interpretation of Rome Ⅲ standard for functional gastrointestinal diseases: functional abdominal pain[J]. J Clin Gastroenterol, 2006, 18(5): 266−267. DOI: 10.3870/j.issn.1005-541X.2006.05.007.[15] Mashayekhi R, Parekh VS, Faghih M, et al. Radiomic features of the pancreas on CT imaging accurately differentiate functional abdominal pain, recurrent acute pancreatitis, and chronic pancreatitis[J]. Eur J Radiol, 2020, 123: 108778. DOI: 10.1016/j.ejrad.2019.108778. [16] Hanania AN, Bantis LE, Feng ZD, et al. Quantitative imaging to evaluate malignant potential of IPMNs[J/OL]. Oncotarget, 2016, 7(52): 85776−85784[2021-11-30]. https://www.oncotarget.com/article/11769/text/. DOI: 10.18632/oncotarget.11769. [17] Eilaghi A, Baig S, Zhang YC, et al. CT texture features are associated with overall survival in pancreatic ductal adenocarcinoma—a quantitative analysis[J/OL]. BMC Med Imaging, 2017, 17(1): 38[2021-11-30]. https://bmcmedimaging.biomedcentral.com/articles/10.1186/s12880-017-0209-5. DOI: 10.1186/s12880-017-0209-5. [18] Chu LC, Park S, Kawamoto S, et al. Utility of CT radiomics features in differentiation of pancreatic ductal adenocarcinoma from normal pancreatic tissue[J]. AJR Am J Roentgenol, 2019, 213(2): 349−357. DOI: 10.2214/AJR.18.20901.