Advanced 3-D ultrasound imaging using matrix probes with integrated microbeamformers

The research in ultrasound echography is currently very active in the development of both new imaging techniques and instrumentation. Special relevance has been recently gained by high-frame-rate (HFR) imaging methods based on the transmission of defocused ultrasound beams rather than focused beams. It has been shown that these methods significantly increase the frame rate, although at the expense of image quality. On the instrumentation side, the research has been significantly boosted by the introduction of open scanners, i.e. scanners that can implement arbitrary transmission sequences and acquire raw echo-data for real-time or offline processing according to novel advanced methods, as programmed by the user. Ultrasound probes have advanced over time: early probes featured a limited number of elements arranged along a line, whereas probes with thousands of elements, arranged in a two-dimensional (2-D) array, have been recently introduced for three-dimensional (3-D) imaging. A key challenge has thus emerged: how to connect and control thousands of elements with the limited number of channels available in ultrasound scanners while still focusing (or beamforming) the energy received by all elements, similarly to a lens. The latest generation of 2-D probes is equipped with the so-called microbeamforming electronics that perform the first step of the beamforming and reduce the number of connections necessary between the probe and the scanner. Unfortunately, such probes are usually compatible only with the scanner produced by the same company that manufactures the probe. Ensuring the compatibility between the new probes and different ultrasound scanners represents a major challenge. The probe electronics must be set up to be controlled by a “generic” scanner, which must offer high flexibility in terms of communication protocols and output interfaces to communicate, configure, and control the probe. This Ph.D. work, conducted in collaboration with Esaote S.p.A. (Genova, Italy), focused on 3-D imaging based on probes embedding a microbeamformer. Both experimental and simulation activities were conducted. The experimental work permitted, for the first time, the connection of a transesophageal echocardiography probe (Oldelft Ultrasound, Delft, The Netherlands) to the 256-channel Ultrasound Advanced Open Platform (ULA-OP 256, MSDLab, University of Florence, Italy). The probe includes a 2-D array with 64×48 elements and microbeamforming electronics that reduce the number of channels to 192, a number compatible with ULA-OP 256. The connection was made successful thanks to the development of additional electronics (to be placed on board ULA-OP 256 as well as on the probe connector), software, and firmware dedicated to the control of the probe and the acquisition of echo signals. Three printed circuit boards were designed and built to enable the above-mentioned connection. The first board allows for an electrical connection between the probe and the ULA-OP 256, the second matches the signal characteristics, and the third one establishes the communication. The functionality of this unique system was evaluated through 3-D imaging and Doppler tests. The experimental results confirm the simulation results. The possibility of working with a microbeamformer-based 2-D probe has also promoted two different simulation studies. The first study was addressed to evaluate the image quality obtainable when such a probe is used in HFR modality. Since the microbeamformers are typically designed for focused beams, their operation is expected to be nonideal when defocused beams are transmitted. Different scan strategies and microbeamformer settings were simulated to determine the optimal compromise between achievable image quality and frame rate. A second simulation study explored the impact of different beamforming algorithms in combination with these novel probes. The filtered delay multiply and sum (FDMAS) algorithm, which has been shown capable of enhancing the image contrast, was implemented in the microbeamformer in combination with the delay and sum (DAS) in the scanner, and vice versa. The study was conducted by varying the beamforming architecture and the microbeamformer settings to identify, for FDMAS, the stage that provides the greatest benefits. During the Ph.D. course, activities collateral to the main one were also carried out. New ultrasound knowledge and technical skills were gained by collaborating on the design of an ultraportable ultrasound system, under development by the MSDLab. Given its portability, compact design, and low cost, this type of system has significant potential as a tool for early diagnosis in critical situations. Furthermore, its flexibility enables the implementation of new imaging techniques and the exploration of new functionalities. A second collateral activity was conducted during a three-month long internship at Technische Universiteit (TU) Delft, The Netherlands, where a new high-frequency probe was under development. The interest in the high-frequency probe stems from the need for high-resolution imaging, here for application to small-animal brain imaging. After the characterization of an existing prototype, the structure of each probe element was modified, and simulations allowed to test the corresponding performance improvement in terms of central frequency, bandwidth, and radiated acoustic pressure.