摘要
目的建立3D CT影像组学模型评估含铂双药联合方案对晚期肺鳞癌的化疗疗效,以期帮助后期治疗方案的选择及调整。方法回顾性分析2019年1月至2021年12月经蚌埠医学院第一附属医院经病理证实的127例晚期肺鳞癌患者的影像学资料,选取化疗前胸部增强CT纵隔窗图像及化疗两个周期后的胸部CT图像,依照实体肿瘤疗效评价标准(RECIST)1.1版评估并分为有效组78例和无效组49例。使用慧影大数据人工智能科研平台将数据按照4∶1随机分为训练组和测试组。选取患者化疗前图像,勾画感兴趣区(ROI),提取病灶的影像组学特征,建立预测模型并验证。通过受试者工作特征曲线(ROC)的曲线下面积(AUC)评估模型效能。结果训练组中,有效组AUC为0.82(0.74~0.91),准确率为0.65,特异度为0.74,敏感度为0.74;无效组AUC为0.82(0.74~0.91),准确率为0.84,特异度为0.77,敏感度为0.77;测试组中,有效组AUC为0.76(0.57~0.96),准确率为0.70,特异度为0.81,敏感度为0.81;无效组AUC为0.76(0.57~0.96),准确率为0.81,特异度为0.70,敏感度为0.70。结论基于治疗前增强CT纵隔窗3D影像组学模型可在化疗前对晚期肺鳞癌的疗效做出较为准确的评估,为患者做出个体化治疗方案具有一定的指导价值。
Objective A 3D CT imaging omics model was established to evaluate the chemotherapy efficacy of platinum-containing two-drug combination regimen in the treatment of advanced lung squamous cell carcinoma,in order to help the selection and adjustment of later treatment regimen.Methods The imaging data of 127 patients with advanced lung squamous cell carcinoma confirmed by pathology in the First Affiliated Hospital of Bengbu Medical College,Anhui Province from January 2019 to December 2021 were retrospectively analyzed.Chest CT images of mediastinal window before chemotherapy and chest CT images of two cycles after chemotherapy were selected.According to the efficacy evaluation criteria for solid tumors(RECIST 1.1),the patients were divided into effective group(78 cases)and ineffective group(49 cases).The data were randomly divided into training group and test group according to 4∶1 by using Hui Ying big data artificial intelligence scientific research platform.The images of patients before chemotherapy were selected to delineate the region of interest(ROI),extract the imaging omics features of lesions,establish and verify the prediction model.The model was evaluated by receiver operating characteristic(ROC)curve.Results In the training group,the AUC of the effective group was 0.82(0.74,0.91),the accuracy was 0.65,the specificity was 0.74 and the sensitivity was 0.74.In the invalid group,AUC was 0.82(0.74,0.91),accuracy was 0.84,specificity was 0.77,sensitivity was 0.77.In the test group,the AUC of the effective group was 0.76(0.77-0.96),the accuracy was 0.70,the specificity was 0.81 and the sensitivity was 0.81.The AUC of the invalid group was 0.76(0.77-0.96),the accuracy was 0.81,the specificity was 0.70,and the sensitivity was 0.70.Conclusion The mediastinal window 3D imaging omics model based on pre-treatment enhanced CT can accurately evaluate the efficacy of advanced lung squamous cell carcinoma before chemotherapy,which has certain guiding value for the individualized treatment of patients.
作者
吴明明
陈艾琪
唐聪聪
杜小萌
卢楚鸣
李想
赵以惠
李伟
左盼莉
马宜传
WU Mingming;CHEN Aiqi;TANG Congcong(Department of Radiology,The First Affiliated Hospital of Bengbu Medical College,Bengbu,Anhui Province 233004,P.R.China)
出处
《临床放射学杂志》
北大核心
2023年第12期1913-1917,共5页
Journal of Clinical Radiology
基金
2022年度研究生科研创新计划项目(编号:Byycx22138)
2021年第二批安徽省中央引导地方科技发展资金项目(编号:2020b07030008)
2021年度研究生科研创新计划项目(编号:Byycxz21042)
领航菁英科研项目专项基金项目(编号:XM-HR-YXFN-2021-05-19)
2023年安徽省临床医学研究转化专项立项项目(编号:202304295107020072)。
关键词
肺鳞癌
增强CT
化疗
3D勾画
影像组学
Lung squamous carcinoma
Contrast-enhanced CT
Chemotherapy
3D sketch Radiomics
