Background:By the definition of the Radiological Society of North America (RSNA) in 1997, tumor ablation is comprised of three parts, which are chemical ablation, thermal ablation and cryoablation. Tumor ablation leads to tissue coagulation necrosis using physical and chemical methods. The thermal ablation has different methods, i.e. radiofrequency ablation, microwave ablation, High intensity focused ultrasound and High Intensity Focused Ultrasound. The success of tumor ablation depends on target location and real-time diagnosis of its biological activity. For more accurate treatment, different kinds of graphic methods are usually used. In our research, we use ultra-sonography, ultrasonic inspection method, CT, Magnetic Resonance Imaging and Magnetic Resonance Imaging guidance methods in clinical. CT has higher image contrast, but it is not designed for dynamic display. Furthermore, it will cause ionizing radiation damage, which restricted its application in practice; On the other hand, MRI has high image resolution, good soft tissue contrast, zero ionizing radiation, non-invasive temperature measurement advantages. But it has poor real-time performance, let alone it is expensive and is sensitive to the variance of temperature. The MRI is difficult to be integrated in ablation device. Compared with CT and MRI, the ultrasound monitoring method has better real-time performance, zero ionizing radiation and also easy to be integrated in the ablation device. So ultrasound is now the most popular image-guided tumor ablation method. The US also has some drawbacks. Currently, ultrasound determines tissue coagulation necrosis by gray-level evaluation, but about 20~30% of coagulation necrosis samples have no obvious change in gray-level. On the other hand, the ultrasonographic speckle statistics characteristics have changed by the ultra-sound treatment. This study attempts to establish bovine liver tissue coagulation necrosis in vitro through HIFU irradiation. Specifically, we intend to improve the determination capability of coagulation necrosis by ultrasonographic similarity analysis.Objective:In this study, real-time monitor bovine liver tissue through HIFU exposuring in vitro by ultrasound ,and then real-time capture sonogram at different time,and then analyse ultrasound speckle image by similarity analysis, and then do histopathology, to explore the following questions: whether the correlation function can be used to determine tissue coagulation necrosis; whether Tamura texture based on wavelet transform analysis can be used to determine the tissue coagulation necrosis; find the the decision hyperplane of tissue coagulation necrosis which detected by correlation function; find the decision hyperplane of tissue coagulation necrosis which detected by Tamura texture based on wavelet transform analysis.Materials and Methods:Freshly bovine liver tissue in vitro, soaked in 0.9% NS, degassing. JC-200 focused ultrasound tumor theraputic system, treatment head frequency is 1.03MHz, irradiation power is 90W,and radiation depth is 20mm. Irradiation at different times (1 s, 2s, 3s ... ... 10s), collected HIFU irradiation before and after at different times (0,1 min, 2 min ... ... 5 min) target two-dimensional ultrasonography. Correlation coefficient analyse over speckle image, and then use SVM(Support Vector Machine) to select parameters and to get the decision hyperplane. After irradiation, cut tissue, showed the largest cross section, measure the maximum damage diameter line, after the digital camera to take normal tissue and necrotic tissue at the junction of the regional organization for pathology examination.Results:1.ultrasound speckle image correlation function can evaluate HIFU coagulation necrosis, the sensitivity was higher than using gray evaluation (Χ2 = 18.716, P <0.05). 2.The area under the ROC curve of ultrasound speckle image correlation function determining bovine liver ablation organizations is between 0.70 and 0.90,and the area is larger than thearea of gray scale evaluation.3.Texture based on wavelet transform analysis ultrasound speckle image can evaluate HIFU Tamura coagulation necrosis, the sensitivity was higher than with the gray evaluation (Χ2 = 15.211, P <0.05).4.The area under the ROC curve of ultrasound speckle images Tamura texture analysis based on wavelet transform Tamura texture analysis determining bovine liver ablation organizations is between 0.70 and 0.90, and the area is larger than thearea of gray scale evaluation.Conclusion:1.sonographic correlation coefficient to judge target tissue coagulation necrosis of HIFU irradiation is better than gray-scale evaluation.2.ultrasonography tamura texture analysis based on wavelet transform analysis to judge target tissue coagulation necrosis of HIFU irradiation is better than gray-scale evaluation.3.similarity analysis over the ultrasound speckle images of coagulation necrosis can improve the capacity of ultrasound images real-time monitoring, and it is expected to become the new HIFU Clinical evaluation of real-time mode. |