Integration of Wave Theory and Statistical Pattern Classification for Ultrasound Imaging of Breast Tumors

在本研究中，将成像处理技术与波动理论相结合，开发一种计算效率高的乳腺小肿瘤超声成像算法。该研究还包括通过基于统计模式分类的材料分类步骤，真实地包含所涉及组织的材料性质和散射中心和结构的真实分布。初步结果表明，该方法具有良好的重建效果，可以区分肿瘤和囊肿。这项研究的最终目标是在实验中强散射环境中检测毫米大小的乳腺肿瘤。该算法对其他存在强散射问题的地区也很有用。

ECS-9711051 Shira Broschat In this research, a computationally efficient algorithm for ultrasound imaging of small breast tumors will be developed by combining imaging processing techniques with wave theory. The research also includes, through its Materials Classifications step based on statistical pattern classification, realistic inclusion both of materials properties of the tissues involved and realistic distribution of scattering centers and structures. Preliminary results indicate that this approach provides excellent reconstruction and can distinguish between tumors and cysts. The ultimate goal of the research is the detection of millimeter-sized breast tumors in what is experimentally a strong scattering environment. The algorithm will also prove useful for other areas where strong scattering is a problem.