Large aperture and wideband modular ultrasound arrays for the diagnosis of liver cancer

我们的项目有两个基本目标：1)开发模块化大口径、高通道数阵列， 相关的电子产品，并使这些模块广泛提供给学术界，以及2)使用这些 改善腹部超声(US)诊断肝癌，特别是难以成像的肝癌的阵列 病人。原发性肝癌患者的总体5年相对存活率为16%。使用了图像处理 监测由病毒感染或肝硬变引起的肝病，并检测向恶性肿瘤的转变，以及 超声波是唯一推荐的筛查此类癌症风险患者的方法。然而，在患有糖尿病的患者中 腹壁厚度大于2.5 cm时，仅有33%的病变被发现。超过三分之二的 美国人现在超重或肥胖，更大的通道数阵列和更大的阵列占用空间将 改善这一人群中的成像，并改善普通人群中小病变的检测。 我们的方法解决了更高的分辨率(大孔径和带宽)、灵敏度(单晶 传感器)、用于超分辨率成像的高帧速率(数百帧/秒可行)、成品率(阵列 由高产量模块组成)和图像对比度(新的切换功能支持新的波束形成 方法)。横向US分辨率与换能器孔径成反比，因此， 提高分辨率需要增加传感器元件和电子通道的数量，以及 先进的波束形成方法。即使在存在异常组织的情况下，通过 扩展的孔径便于对小结构的可视化。我们建议创建PIN-PMN-PT阵列 尺寸为16个元素(方位角)乘32个元素(仰角)的模块，它们将组合在一起形成大型 数组。用户可选择的ASIC矩阵提供了从单个 模块，用于在高程上组合镜像元素或在方位角上组合相邻元素 立面。结果，大口径阵列内的4096个单元可以同时从 可编程扫描仪。初步工作已经完成了原型阵列和ASIC模块的制造。 加州大学戴维斯分校的团队是首批生产多频率阵列的团队之一；南加州大学率先推出了高频率阵列 频率阵列，目前是Wodicki的总部(在超声ASIC开发方面有20年的经验 在GE)，并将与我们合作开发高通道数阵列；杜克大学 率先开发使用大孔径的策略；Verasonics是领先的 并将与我们合作开发数据处理方法并分发 超声波社区的组件；Sonic Concepts将制造结果阵列；GE有 作为外部顾问在将这些阵列整合到其他商业系统方面提供支持。 这个独特的团队将开发这项技术，评估其在人体研究中的使用，并将传播 技术

Our project has two fundamental goals: 1) develop modular large aperture, high channel count arrays with associated electronics and make these modules widely available to the academic community, and 2) use these arrays to improve abdominal ultrasound (US) for the diagnosis of liver cancer, particularly for difficult to image patients. The overall 5-year relative survival rate for patients with primary liver cancer is 16%.  Imaging is used to monitor liver disease resulting from viral infections or cirrhosis and to detect a transition to malignancy, and ultrasound is the only recommended method for screening such patients at risk for cancer.  Yet, in patients with an abdominal wall thickness greater than 2.5 cm, only 33% of lesions were detected. More than 2/3 of Americans are now overweight or obese and larger channel count arrays and larger array footprints will improve imaging within this population and improve the detection of small lesions in the general population. Our approach addresses improved resolution (large aperture and bandwidth), sensitivity (single crystal transducers), high frame rate for super-resolution imaging (hundreds of frame/sec feasible), yield (array is composed of high yield modules), and image contrast (new switching capabilities enable new beamformation methods). Lateral US resolution is inversely proportional to the transducer aperture and consequently, improved resolution requires an increased number of transducer elements and electronic channels as well as advanced beam formation methods. Even in the presence of aberrating tissues the contrast achieved from an extended aperture facilitates the visualization of small structures. We propose to create PIN-PMN-PT array modules of dimension 16 elements (azimuth) by 32 elements (elevation) which will be combined to form large arrays. The user selectable ASIC matrices provide the opportunity to select: 512 elements from within a single module, to combine mirrored elements in elevation or to combine neighboring elements in azimuth or elevation. As a result, 4096 elements within the large aperture array can simultaneously be addressed from a programmable scanner. Preliminary work has resulted in the fabrication of prototype array and ASIC modules. The UC Davis team has been one of the first to produce multi-frequency arrays; USC has pioneered high frequency arrays and is currently home to Wodnicki (20 years of experience in ultrasound ASIC development at GE) and will collaborate with us on the development of high channel count arrays; Duke University has pioneered the development of strategies to use large apertures; Verasonics is the leading manufacturer of programmable ultrasound systems and will work with us to develop data handling methods and to distribute components to the ultrasound community; Sonic Concepts will manufacture the resulting arrays; and GE has offered support as an external advisor on the incorporation of these arrays into other commercial systems. This unique team will develop the technology, evaluate its use in a human study and will disseminate the technology.