Compressive sensing-based sparse transducers for ultrasound imaging

超声成像通常用于材料测试和许多疾病的诊断。超声成像的分辨率由各种因素决定，包括换能器设计和工作频率。为了最大限度地提高分辨率，通常采用高工作频率（> 1 MHz）和多元件换能器，这导致两个主要限制：1）穿透深度受到超声波在传播介质中的大衰减以及与高频下的界面相关联的畸变和混响的限制，2）与大量换能器元件相关联的信号处理的复杂性仍然限制了对于给定图像质量可实现的帧速率。到目前为止，所进行的大部分工作都集中在成像算法或换能器设计上。独特的使能技术组合将使超声成像的穿透深度和帧速率实现重大突破：1）成像算法Excitelet，其利用较低频率超声并且固有地考虑全波传播路径的详细模型，2）压缩感测算法，其利用编码传播路径中的多样性，以及3）增材制造技术，其使得能够对超材料内的波传播路径进行编码。该计划的目标是创建新一代稀疏超声换能器阵列，其元件数量少，信号处理复杂性低，帧速率高，工作频率低，穿透深度高，因此打破了现有的方法。该阵列将用于介质的超声成像，该介质可以是工程材料或生物组织。在大穿透深度和1-10 kHz量级的帧速率下，寻求低至波长的1/30的横向分辨率。这项研究计划阐明了沿着三个研究方向，平行但相关：1）利用超声换能器的压缩感测，其中波和频率的类型将用于丰富信息编码，2）稀疏换能器阵列的优化，其中Excitelet将在优化过程中被考虑，以及3）超声探头制造中的超材料，其中将使用增材制造技术来制造用于编码探针的信息和其它组件的超材料。在材料测试中，这项工作将允许更好地表征先进材料中的损伤前兆，例如复合材料结构（微裂纹，空隙，微屈曲，局部纤维断裂或局部纤维基质脱粘）。在医学诊断中，这项工作将有助于满足对快速成像帧速率和近场和远场更好分辨率的迫切需求，以避免引导程序中周围器官的风险，此外还可以扩展穿透深度，以适应肥胖人口比例的增加。

Ultrasound imaging is used routinely for material testing and for the diagnosis of many diseases. The resolution of ultrasound imaging is determined by various factors, including transducer design and frequency of operation. To maximize resolution, high operating frequencies (> 1 MHz) and multi-element transducers are usually employed, leading to two major limitations: 1) the penetration depth is limited by the large attenuation of the ultrasonic waves in the propagating medium and by the aberration and reverberation associated with the interfaces at high frequencies, 2) the complexity of the signal processing associated with a large number of transducer elements still limits the achievable frame rate for a given image quality. Until now, most of the work conducted has focused on either imaging algorithms or transducers design. A unique combination of enabling technologies will allow a major breakthrough in the penetration depth and frame rate in ultrasound imaging: 1) the imaging algorithm Excitelet, which exploits lower frequency ultrasound and inherently considers a detailed model of the full wave propagation path, 2) compressive sensing algorithms, which exploit diversity in encoded propagation paths, and 3) additive manufacturing technologies, which enable encoding wave propagation paths within metamaterials. The objective of this program is to create a new generation of sparse ultrasound transducer arrays with a low number of elements, for lower signal processing complexity and higher frame rate, and working at low frequency for improved penetration depth, therefore breaking with existing approaches. The arrays will be used for ultrasound imaging of media which can be either engineering materials or biological tissues. Lateral resolution down to 1/30 of the wavelength is sought at large penetration depth and at a frame rate in the order of 1-10 kHz. This research program articulates along three research thrusts, parallel but related: 1) compressive sensing with ultrasound transducers, where the type of waves and frequencies will be used to enrich information encoding, 2) optimization of sparse transducer arrays, where Excitelet will be considered within the optimization process, and 3) metamaterials in ultrasound probe fabrication, where additive manufacturing techniques will be used to fabricate the metamaterials for encoding the information and other components of the probe. In material testing, this work will allow better characterization of damage precursor in advanced materials, such as composite structures (micro-cracks, voids, micro-buckling, local fiber breakage, or local fiber-matrix debonding). In medical diagnosis, this work will contribute to answer an urgent need for fast imaging frame rates and better resolution in both near- and far-field to avoid the risks in the surrounding organs in guided procedures, in addition to extended penetration depth in order to adapt to the increasing proportion of the obese population.