Development of Three Dimensional Tissue Elasticity Imaging -Acoustic Palpation- System based on Combined Autocorrelation

组织弹性性质的变化与组织的病理状态密切相关。因此，静态压缩时的应变分布图像虽然比弹性模数分布更定性，但更容易检测，但对于基于组织弹性特性差异的疾病组织检测是有用的。然而，大多数方法都是基于空间相关技术，计算量大，不适合实时处理。虽然另一种方法，即相位跟踪法可以实现应变的快速检测，但由于应变图像的高信噪比，它不能工作大的位移。因此，从可行性的角度出发，我们开发了一种新的组织弹性成像技术。提出的方法称为组合自相关方法，具有ob…的优点。另一方面，为了获得更多的组织的定量特征，研究了一些弹性成像的方法。然而，为了简化估算问题，许多估算弹性模量分布的方法都是基于二维模型，例如平面应力分布或平面应变分布，这对于实际组织来说并不总是能满足要求。因此，我们发展了一种基于三维组织模型的更实际条件下的组织弹性模数估计方法，数值模拟和体模实验表明，与传统的空间相关方法相比，该方法在实际条件下具有更高的重建组织应变分布图像的能力。此外，通过三维组织模型估计应力分布，还得到了弹性模量场的图像。利用该方法对乳腺肿瘤标本进行了测量，使得得到的应变图像能够清晰地检测出B超图像中模糊的较硬的肿瘤病变。此外，实验结果还表明，提取肿瘤的病理特征有助于肿瘤类型的鉴别。验证了该方法在临床应用的可行性。较少

The variation of elastic properties of tissues have much relation to the pathological condition. Therefore, the image of strain distribution when static compression is applied are proven to be useful for detecting the disease tissue based on the difference of tissue elastic properties though it is more qualitative but easier to detect than elastic modulus distribution.So far, many approaches to obtain strain image from echo signals have been proposed. However, most methods are based on the spatial correlation technique and do not suit for real-time processing since they require much computation time. Although alternative approach, that is, a phase tracking method can realize the rapid detection of strain, it does not work a large displacement required for the high SN ratio of strain image. Therefore, we developed a new technique for tissue elasticity imaging from the viewpoint of the feasibility. The proposed approach referred to as the combined autocorrelation method has a merit of ob … More taining high quality strain image with real-time processing and being applicable to a large displacement.On the other hand, some approach for elasticity imaging are investigated for the purpose of obtaining more quantitative properties of tissues. However, in order to simplify the estimation problem, many approaches for estimating elastic modulus distribution.are based on 2-D model, for example, plane stress distribution or plane strain distribution, which is not always satisfied for actual tissues. Therefore, we have developed a method of estimating the elastic modulus under more practical conditon based on a 3-D tissue model.Numeric simulation and phantom experiment demonstrated that the method has a higher capability of reconstructing the image of tissue strain distribution in practical condition as compared with conventional spatial correlation method. Moreover, image of elastic modulus distribution was also obtained by estimating stress distribution using 3-D.tissue model. Measurement of breast tumor specimen is performed by the proposed method so that the obtained strain image clearly detected the harder tumor lesions which were vague in B-mode image. Moreover, the results indicated the possibility of extracting the pathological characteristic of tumor to be useful for discrimination of tumor type. These validate the feasibility of the proposed method for clinical application. Less