论文部分内容阅读
目的筛选骨折影像学检验CT片上的观察指标,建立新旧眶内壁骨折的鉴别方法。方法收集70例新鲜和51例陈旧性爆裂性眶内壁骨折CT影像学资料,筛选出与新旧骨折相关的8项指标,使用SPSS 13.0软件对各指标进行分析,建立单一指标判别方程;根据分析结果,剔除准确率低的观察指标。再使用SPSS 13.0软件逻辑回归分析的方法,建立多指标综合判别方程,并进行准确性验证。结果各单一指标方程验证结果显示,综合判别准确率最高为95.9%,最低为57.9%,剔除准确率最低的两个指标后建立两个多指标判别方程,经验证,方程(1)判别准确率为99.2%(χ2=124.382,P=0.000),方程(2)判别准确率为95.0%(χ2=159.339,P=0.000)。结论利用CT片上的形态学指标,建立相关判别方程式,可有效鉴别新旧爆裂性眶壁骨折。联合使用多指标的判别方程,结果更准确。
Objective To screen the observation indexes of CT images of fracture imaging and establish the identification method of old and new orbital wall fractures. Methods Seventy-eight cases of fresh or 51 cases of old orbital orbital wall fractures were collected. Eight indicators related to old and new fractures were screened. SPSS 13.0 software was used to analyze each index to establish a single index discriminant equation. According to the analysis results , Remove the low accuracy of the observation. Then using the method of SPSS 13.0 software logistic regression analysis, the multi-index comprehensive discriminant equation is established and the accuracy is verified. Results The results of the single index equation validation showed that the accuracy of comprehensive discrimination was 95.9% and the lowest was 57.9%. Two multi-criteria discriminant equations were established after removing the two indicators with the lowest accuracy rate. It was verified that the discriminant accuracy of equation (1) Was 99.2% (χ2 = 124.382, P = 0.000). The discriminant accuracy of equation (2) was 95.0% (χ2 = 159.339, P = 0.000). Conclusion Based on the morphological indexes of CT, the discriminant equation can be established to discriminate the old and new orbital fracture. The joint use of multiple indicators of the discriminant equation, the result is more accurate.