High-intensity focused ultrasound (HIFU) multiple lesion imaging: comparison of detection algorithms for real-time treatment control
High-intensity focused ultrasound (HIFU) multiple lesion imaging: comparison of detection algorithms for real-time treatment control
复制标题
DOI：
10.1109/ultsym.2002.1192564
复制
发表时间：
2002-10
期刊：
影响因子：
--
通讯作者：
R. Seip;J. Tavakkoli;R. Carlson;A. Wunderlich;N. Sanghvi;K. Dines;T. Gardner
中科院分区：
文献类型：
--
作者：
R. Seip;J. Tavakkoli;R. Carlson;A. Wunderlich;N. Sanghvi;K. Dines;T. Gardner
HIFU引起的病变的成像提供了非侵入性的实时治疗监测和控制。这项工作提出了专门为多个病变检测设计的HIFU诱导的病变检测算法所获得的结果。比较了对HIFU期间相对组织变化敏感的算法-测量信号能量、组织位移、熵和组织衰减，以确定其检测多个和相邻HIFU损伤的创建的能力。在7次治疗期间，使用定制Sonablate/spl reg/500 HIFU器械采集体内（N=4）犬前列腺背散射RF数据。总共治疗了815个部位，形成了算法评估数据集。结果发现，基于信号能量的算法表现最好，检测到82%的所有HIFU损伤，同时显示低于5%的误报率。所有方法都是完全非侵入性的，并且利用在HIFU治疗之前、期间和之后获得的组织参考/标准化信息。算法细节，数据采集方法，在体内的实验结果，和算法比较结果。
Imaging of HIFU-induced lesions provides non-invasive, real-time treatment monitoring and control. This work presents results obtained with HIFU-induced lesion detection algorithms specifically designed for multiple lesion detection. Algorithms sensitive to relative tissue changes during HIFU -measuring signal energy, tissue displacement, entropy, and tissue attenuation are compared for their ability to detect the creation of multiple and adjacent HIFU lesions. In vivo (N=4) canine prostate backscattered RF data was acquired with a custom Sonablate/spl reg/500 HIFU device during 7 treatments. A total of 815 sites were treated, forming the algorithm evaluation dataset. It was found that the algorithm based on signal energy performed best, detecting 82% of all HIFU lesions created, while showing false-alarm rates below 5%. All methods are completely non-invasive, and make use of tissue reference/normalization information obtained before, during, and after the HIFU treatment. Algorithm specifics, data acquisition methodologies, in vivo experimental results, and algorithm comparison results are shown.