| 姬慧君,宁尚昆,刘浅浅,陈盈秀,陈琪,顾梦瑶,黄勇,李万湖.平扫CT影像组学鉴别肺炎型黏液腺癌与大叶性肺炎[J].中国医学影像技术,2023,39(4):549~554 |
| 平扫CT影像组学鉴别肺炎型黏液腺癌与大叶性肺炎 |
| Plain CT radiomics for differentiating pneumonia type mucinous adenocarcinoma and lobar pneumonia |
| 投稿时间：2022-11-07 修订日期：2023-01-08 |
| DOI：10.13929/j.issn.1003-3289.2023.04.014 |
| 中文关键词: 肺肿瘤 肺炎 影像组学 体层摄影术,X线计算机 |
| 英文关键词:lung neoplasms pneumonia radiomics tomography, X-ray computed |
| 基金项目: |
|
| 摘要点击次数: 1595 |
| 全文下载次数: 760 |
| 中文摘要: |
| 目的 观察平扫CT影像组学鉴别肺炎型黏液腺癌(PTMA)与大叶性肺炎(LP)的价值。方法 回顾性分析57例PTMA(PTMA组)和129例LP患者(LP组),按7 ∶ 3比例将其纳入训练集(n=131)和测试集(n=55)。比较组间患者临床资料,筛选临床特征,构建临床模型;勾画病灶ROI,提取其影像组学特征,建立影像组学模型;基于临床特征及影像组学特征建立列线图模型。观察3种模型鉴别PTMA与LP的效能。结果 组间患者年龄和呼吸道症状占比差异均有统计学意义(P均＜0.05)。临床模型鉴别训练集和测试集PTMA与LP的曲线下面积(AUC)分别为0.784和0.909。最终纳入16个影像组学特征建立影像组学模型,其在训练集和测试集鉴别PTMA与LP的AUC分别为0.909和0.870;列线图模型的AUC分别为0.939和0.933。影像组学模型及列线图模型在训练集鉴别PTMA与LP的AUC均大于临床模型(P均＜0.05)。结论 平扫CT影像组学有助于鉴别PTMA与LP。 |
| 英文摘要: |
| Objective To observe the value of plain CT radiomics for differentiating pneumonia type mucinous adenocarcinoma (PTMA) and lobar pneumonia (LP). Methods Fifty-seven patients with PTMA (PTMA group) and 129 LP patients (LP group) were enrolled and divided into training set (n=131) and test set (n=55) at the ratio of 7 ∶ 3. Patients' clinical data were compared between groups, and clinical features were screened to establish a clinical model. ROI of lesions were delineated on chest CT images, then the radiomics features of lesions were extracted to establish a radiomics model. Finally a nomogram model was established based on combination of clinical features and radiomics features. The efficacy of each model for differentiating PTMA and LP was evaluated. Results Significant differences of patients' age and respiratory symptoms were found between PTMA group and LP group (both P＜0.05). The area under the curve (AUC) of clinical model for differentiating PTMA and LP in training set and test set was 0.784 and 0.909, respectively. Totally 16 radiomics features were enrolled. The AUC of radiomics model for differentiating PTMA and LP and in training set and test set was 0.909 and 0.870, respectively, while of nomogram model was 0.939 and 0.933, respectively. In training set, AUC of radiomics model and nomogram model were both higher than that of clinical model (both P＜0.05). Conclusion Plain CT radiomics was helpful to differentiating PTMA and LP. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
