Using passive cavitation images to classify high-intensity focused ultrasound lesions.
Using passive cavitation images to classify high-intensity focused ultrasound lesions.
复制标题
DOI：
10.1016/j.ultrasmedbio.2015.04.025
复制
发表时间：
2015-09
影响因子：
2.9
通讯作者：
Mast TD
中科院分区：
文献类型：
--
作者：
Haworth KJ;Salgaonkar VA;Corregan NM;Holland CK;Mast TD
被动空化成像提供了对空化排放的空间分辨监测。然而，线性成像阵列的衍射极限导致相对较差的距离分辨率。差的距离分辨率限制了用于预测热损伤形成的被动空化成像的空间特异性和灵敏度的先前分析。在这项研究中，这一限制是克服定向的线性阵列正交的HIFU传播方向和执行被动成像。1.1 MHz连续波超声暴露后，离体牛肝脏样本中形成了14个病变。病变根据其形状和位置分为局灶性、“蝌蚪状”或局灶前病变。被动空化图像波束形成的基础上，谐波，超谐波和非谐波频率的发射与既定的算法。使用受试者工作特征曲线（AUROC）下的面积，基波，谐波和超谐波发射被证明是所有病变类型的病变形成的重要预测因子。对于谐波和超谐波发射，局灶前病变的分类最成功（AUROC值分别为0.87和0.88），其次是蝌蚪病变（AUROC值分别为0.77和0.64）和局灶性病变（AUROC值分别为0.65和0.60）。
Passive cavitation imaging provides spatially resolved monitoring of cavitation emissions. However the diffraction limit of a linear imaging array results in relatively poor range resolution. Poor range resolution has limited prior analyses of the spatial specificity and sensitivity of passive cavitation imaging for predicting thermal lesion formation. In this study, this limitation is overcome by orienting a linear array orthogonal to the HIFU propagation direction and performing passive imaging. Fourteen lesions were formed in ex vivo bovine liver samples as a result of 1.1 MHz continuous-wave ultrasound exposure. The lesions were classified as focal, “tadpole”, or pre-focal based on their shape and location. Passive cavitation images were beam-formed from emissions at the fundamental, harmonic, ultraharmonic, and inharmonic frequencies with an established algorithm. Using the area under a receiver operator characteristic curve (AUROC), fundamental, harmonic, and ultraharmonic emissions were shown to be significant predictors of lesion formation for all lesion types. For both harmonic and ultraharmonic emissions, pre-focal lesions were classified most successfully (AUROC values of 0.87 and 0.88, respectively), followed by tadpole lesions (AUROC values of 0.77 and 0.64, respectively), and focal lesions (AUROC values of 0.65 and 0.60, respectively).