摘要
目的:基于CT灌注成像动脉期图像探讨纹理分析对胆源性急性胰腺炎(BAP)和高脂血症性急性胰腺炎(HLAP)的鉴别价值。方法:回顾性连续收集2018年4月至2020年9月本院及德阳市人民医院收治的80例首次诊断为急性胰腺炎患者的临床资料及CT灌注影像资料,其中BAP患者29例,HLAP患者51例。将胰腺实质显示最佳的动脉期图像进行5 mm的厚层重组。应用3D Slicer软件逐层手动勾画感兴趣区(ROI),并进行三维融合获得胰腺容积感兴趣区(VOI)后提取纹理特征。利用单因素分析及最小绝对收缩和选择算子(LASSO)回归进行纹理特征选择。收集性别、年龄、实验室检查数据等临床资料进行多因素分析,并建立一个临床模型,选择最佳的纹理特征联合临床资料建立纹理模型。受试者工作特征(ROC)曲线评估两个模型的鉴别效能。结果:临床模型的ROC曲线下面积(AUC)为0.911 (95%CI 0.848~0.974),纹理模型的AUC为0.976 (95%CI 0.951~1.000)。纹理模型的诊断效能明显优于临床模型(P=0.011)。结论:基于胰腺动脉期的纹理模型在鉴别BAP和HLAP方面具有较高的诊断效能,早期鉴别诊断有助于临床的对因治疗。
Purpose: To explore the value of texture analysis of CT perfusion imaging in differentiating biliary acute pancreatitis(BAP) from hyperlipidemic acute pancreatitis(HLAP). Methods: Clinical and CT perfusion imaging data of 80 patients with acute pancreatitis diagnosed for the first time from April 2018 to September 2020were enrolled, including 29 patients with BAP and 51 patients with HLAP. The best arterial phase images of pancreatic parenchyma were reconstructed by 5 mm thickness. The 3D Slicer software was used to manually outline the region of interest(ROI), and conduct three-dimensional fusion to obtain the volume of interest(VOI) of pancreas to extract texture features. And least absolute shrinkage and selection operator(LASSO) regression was used to select texture features. Gender, age, laboratory data and other clinical data were collected for multi-factor analysis to establish a clinical model. The best texture features were selected and combined with clinical data to establish a texture model. The identification efficiency of the two models was evaluated by receiver operating characteristic(ROC) curve.Results: The area under the ROC curve(AUC) of the clinical model was 0.911(95%CI 0.848-0.974),and the AUC of the texture model was 0.976(95%CI 0.951-1.000). The diagnostic efficiency of texture model was significantly higher than that of clinical model(P=0.011). Conclusion: Texture model has high efficiency in the differentiation of HLAP and BAP, which can provide reference for the early clinical symptomatic treatment.
作者
贾清
黄小华
明兵
唐玲玲
胡云涛
刘琢玉
JIA Qing;HUANG Xiaohua;MING Bing;TANG Lingling;HU Yuntao;LIU Zhuoyu(Departmentof Radiology,The Affiliated Hospital of North Sichuan Medical College;Department of Radiology,Deyang People's Hospital)
出处
《中国医学计算机成像杂志》
CSCD
北大核心
2022年第6期603-608,共6页
Chinese Computed Medical Imaging
基金
南充市市校合作科研专项-重要研发平台(19SXHZ0429)
南充市市校科技战略合作(20SXQT0303)。
