| 韩树高,孙浩然,白人驹,汪俊萍,李亚军.MSCT对急性阑尾炎病理类型的预测价值[J].中国医学影像技术,2012,28(5):943~947 |
| MSCT对急性阑尾炎病理类型的预测价值 |
| Predictive value of MSCT for pathological type of acute appendicitis |
| 投稿时间：2011-09-01 修订日期：2011-10-17 |
| DOI： |
| 中文关键词: 阑尾炎 体层摄影术,X线计算机 病理学 |
| 英文关键词:Appendicitis Tomography, X-ray computed Pathology |
| 基金项目: |
|
| 摘要点击次数: 2721 |
| 全文下载次数: 1114 |
| 中文摘要: |
| 目的 评价MSCT对急性阑尾炎病理类型的预测价值。方法 回顾性分析112例经手术病理证实的急性阑尾炎患者的MSCT图像,观察急性阑尾炎的CT征象并进行评分;使用Spearman秩相关法对CT征象与病理分型进行相关性分析,筛选出与病理分型显著相关的CT征象,再利用上述征象作为变量进行Logistic回归分析,以预测病理类型,并评价回归模型的预测效果。结果 阑尾直径、阑尾石、周围脂肪密度、管腔外积液、管腔内积液、箭头征、右侧腰大肌前缘模糊、盆腔积液、管腔内积气、盲肠条带征、乙状结肠壁增厚11个征象与急性阑尾炎病理分型存在显著相关,Logistic回归分析显示前5个征象可预测阑尾炎病理类型,预测值和病理分型的加权κ值为0.78,Logistic回归模型总体诊断准确率为85.71%(96/112)。结论 应用上述5个MSCT征象建立的Logistic回归模型能够准确预测急性阑尾炎患者的病理学类型。 |
| 英文摘要: |
| Objective To evaluate the predictive value of MSCT in predicating pathological type of acute appendicitis. Methods MSCT images of 112 patients with acute appendicitis confirmed pathologically were retrospectively reviewed. MSCT findings were cataloged and scored by standardized criteria. Spearman rank correlation method was performed to screen the MSCT findings which were significantly correlated with pathological types. An ordinal Logistic regression model was established with those CT findings which were tested to be statistically correlated with pathological types. Then the predicted severity values were generated from the model and compared with pathological results. Results There were 11 MSCT findings that statistically correlated with pathological types of acute appendicitis, including appendix diameter, appendolithiasis, density of peripheral fat, extraluminal fluid collection, intraluminal fluid collection, arrowhead sign, right musculi psoas major leading fuzzy, fluid of pelvic cavity, intraluminal gas, cecal bar sign and sigmoid colonic wall thickening. And the former 5 MSCT findings were proved to be able to predict pathological types of acute appendicitis. The weighted κ measurement of agreement between predicted and pathological type was 0.78, while the accuracy of predicting pathological type was 85.71% (96/112). Conclusion An ordinal Logistic regression model with 5 MSCT findings mentioned above as variables is able to accurately predict the pathological type of acute appendicitis. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
