首页 | 本学科首页   官方微博 | 高级检索  
     

医学图像处理技术在乳腺弹性成像技术中的应用价值
引用本文:叶磊,邓克学,隋秀芳,康冰飞,王磊,张杰,崔亚云. 医学图像处理技术在乳腺弹性成像技术中的应用价值[J]. 安徽医科大学学报, 2017, 52(9). DOI: 10.19405/j.cnki.issn1000-1492.2017.09.021
作者姓名:叶磊  邓克学  隋秀芳  康冰飞  王磊  张杰  崔亚云
作者单位:安徽医科大学附属省立医院超声科,合肥,230001;安徽医科大学附属省立医院影像科,合肥,230001
基金项目:安徽省科技计划项目,安徽高校省级自然科学研究项目
摘    要:
目的 研究医学图像处理技术在乳腺弹性成像鉴别肿块良恶性中的价值.方法 收集122例乳腺结节患者,共124个病灶,均接受常规超声、声脉冲辐射力成像技术检查声触诊组织成像(vTI).对VTI图像进行图像处理分析,通过Image J软件测得图像平均光密度值.以病理为金标准,比较每种方法及联合后的诊断效能.结果 经病理证实,124例乳腺肿块中,恶性病灶50例,良性病灶74例.VTI的平均光密度值的截点值为210.200,AUC为0.873.恶性病灶平均光密度值>210.200的患者为45例,<210.200的患者为5例,良性病变平均光密度值>210.200的患者为7例,<210.200的患者为67例(P <0.005).两者联合方法在诊断乳腺肿瘤的敏感度为96.0%、特异度为83.8%、阳性预测值为80.0%、阴性预测值为96.9%、准确率为88.7%.结论 乳腺肿块VTI成像后,医学图像处理技术可以通过平均光密度这个量化的指标来评估组织的硬度,在预测乳腺病变良恶性方面有重要意义,其与常规超声联合判断,并且能明显提高诊断的准确性.

关 键 词:超声弹性成像  乳腺  医学图像  平均光密度

The value of medical image processing in elastography of breast
Abstract:
Objective To explore the value of medcial image processing in the differential of benign and malignant breast masses with elastograph.Methods 124 cases of 122 patients were given common ultrasonic and VTI examination.Analyzing the VTI images with image processing,and acquiring the average optical density.Diagnosis efficiency of each method and combined diagnosis were comparied on the basis of the pathology.Results 124 cases included 74 benigens and 50 malignants.The cut-off value of the average optical density was 210.200 and AUC was 0.873.There were 45 cases > 210.200,and 5 cases < 210.200 in all malignants.There were 7 cases >210.200,67 cases < 210.200 in all benigns (P < 0.05).The sensitivity,specificity,positive,predictive value,negative value and accuracy of combined ultrasound were 96%,83.8%,80%,96.9%,88.7%.Conclusion The hardness of breast mass can be evaluated with average optical density.There was important significance in predicting benign and malignant of breast masses.If combining with common ultrasound,it can improve the diagnostic accuracy.
Keywords:ultrsound  breast  medcial image  optical density
本文献已被 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号