Objective To evaluate the value of GE′s Q-analysis real-time ultrasonic elastography quantitative analysis software in the diagnosis of breast diseases and explore the best cutoff point of the breast diseases diagnosis.Methods The elastograms of 71 breast lesions(66 patients)were evaluated by using the GE LOGIQ E9 with the Q-analysis software.All the 71 lesions were diagnosed as malignant and benign lesions by surgery or biopsy.The receiver operating characteristic (ROC)curves of the entirety elasticity rate and local elasticity rate was performed.We compared the areas under ROC curves ( AUC) of the two modalities and confirmed the best cutoff point of the breast diseases diagnosis .We also assessed the sensitivity and specificity of the two modalities .Lastly we assessed the ROC curves with Z test to determine whether the difference between the two modalities was significant .Results Among the 71 leisions (66 patients),there were 37 benign lesions and 34 malignant ones.The elastograms of the leisions showed that the elasticity of 37 benign leisions was good ,the hardness value was low and the texture was even .As for the 34 malignant leisions,the elasticity was bad,the hardness value was high and the texture was uneven .The AUC of the local modality was 0.899 ( >0.5).The AUC of the entire modality,was 0.985 ( >0.5).Both modalities were accurate for the diagnosis of breast diseases .When the cutoff point of entirety elasticity rate was set as 3.60,the sensitiveness was 82.4%and the specificity was 86.5%.When the cutoff point of local elasticity rate was set as 2.60,the sensitiveness was 97.1% and the specificity was 91.9%.The difference between the two modalities was significant (Z =2.621,P=0.0088).Conclusions Q-analysis real-time ultrasonic elastography quantitative analysis is able to evaluate the rigidity and texture of tissue and is valuable to distinguish benign breast lesions between malignant breast lesions .The sensitiveness and specificity of local modality are higher than that of entirety modality .% 目的评估Q－analysis实时组织弹性成像分析诊断乳腺良、恶性病变的最佳诊断临界值及其在乳腺疾病诊断中的应用价值。方法应用GE LOGIQ E9超声诊断仪及Q－analysis分析软件对66例乳腺肿物患者71个病灶进行超声检查并进一步行实时组织弹性定量分析，所有病例均经手术或穿刺活检获得明确病理结果。绘制乳腺病灶整体弹性比率值与局部弹性比率值的受试者操作特性（ROC）曲线，比较各种病灶整体弹性比率值与局部弹性比率值曲线下的面积，确定适于诊断的临界值，分析两种方法评估乳腺病灶良恶性的敏感度及特异度，并应用Z检验分析所得出的ROC曲线，比较两种方法在乳腺疾病诊断中的差异。结果66例患者71个病灶中良性病灶37个，恶性病灶34个。弹性成像显示37个良性病灶弹性较好，整体硬度值较小，病灶质地较均匀；34例恶性病灶弹性较差，整体硬度值较大，病灶质地不均匀；实时组织弹性分析乳腺病灶整体弹性比率值法与局部弹性比率值法的ROC曲线下面积分别为0．899和0．985，均大于0．5，两种方法对乳腺疾病良恶性的鉴别具有较高的准确性；乳腺病灶整体弹性比率临界值为3．60时，其诊断的敏感度为82．4％，特异度为86．5％，乳腺病灶局部弹性比率临界值为2．60时，其诊断的敏感度为97．1％，特异度为91．9％；两种方法ROC曲线下面积间差异有统计学意义（Z＝2．621，P＝0．0088）。结论Q－analysis实时组织弹性分析技术可用来评估病灶的硬度及均匀程度，为判断乳腺疾病的良恶性提供诊断的依据。整体弹性比率值法与局部弹性比率值法均具有较高的准确性，其中局部弹性比率值法的敏感度及特异度优于整体弹性比率值法。
展开▼
