摘要
目的:统计评价并筛选出鉴别良恶性胃肠道间质瘤(GIST)的超声特征指标。方法:观察71例GIST(良性32例、恶性39例)的术前超声灰阶和彩色多普勒的特征,与恶性GIST有关的变量进入Logistic回归分析,筛选出鉴别GIST良恶性有意义的超声指标,建立诊断恶性GIST可能性的公式。结果:胃肠道良、恶性间质瘤在肿块体积、内部液性回声、形态规则、腹腔淋巴结肿大方面均存在显著性差异(均P<0.05),肿瘤内彩色血流分布无显著性差异(P>0.05)。通过Logistic逐步回归分析,得出方程为:GIST恶性可能性=1/(1+e-z),其中z=-4.248+2.267×肿块体积+3.637×内部液性无回声+3.256×形态不规则+3.071×腹腔肿大淋巴结。当GIST体积大(≥5cm)、肿块内部见不规则液性无回声、形态不规则、伴腹腔淋巴结肿大时,肿瘤的恶性危险可能性为0.88。结论:应用超声特征诊断GIST的Logistic回归模型,可为鉴别肿瘤的良恶性做参考。
Objective:To evaluate statistically the specific characters and index of gastrointestinal stromal tumor(GIST)on gray scale and color Dopper Ultrasound.Methods:The characteristics of the gray scale and color Dopper Ultrasound of the GIST (benign 32 cases,malignant 39 cases)were observed and studied before surgery,and the variables related to the malignant GIST were analyzed via multivariate logistic regression in order to screen out the GIST ultrasonic malignant indexes for identification and establish a diagnosis formula of malignant GIST on the probability.Results:The tumor size,irregnlar shape,swelling lymph nodes and inter- nal echo level were related to its pathologic risk (P〈0.05).There was no statistical relationship between tumor blood flow distribution and malignant GIST risk(P〉0.05).Afier logistic regression analysis,probability of malig- nancy=l/(l+e-2),of which z=-4.248+2.267×tnmor size+3.637×internal echo level+3.256×irregular shape+3.071× swelling lymph nodes.Conclusion:Uhrasonography plays an important role in judging the resource of GIST.The logistic regression model can be refered as a way of differential diagnosis.
出处
《中日友好医院学报》
2015年第5期281-283,共3页
Journal of China-Japan Friendship Hospital
关键词
胃肠间质瘤
超声
鉴别诊断
价值
gastrointestinal stromal tumor
uhrasonography
differential diagnosis