Micromachined 2D Transducers for 3-D Super-Resolution Ultrasound Real-Time Imaging of Erythrocytes

2D ultrasound is a conventional medical diagnostic tool preferred for its real-time, noninvasive and radiation-free imaging technique at low cost compared to other imaging modalities.
Since human anatomy is inherently 3D, reducing it to 2D will inevitably be at a cost of insufficient understanding of the complexity of the body and its diseases. Naturally, the interest in 3D ultrasound has risen. The transducer complexity increases immensely when transitioning from 2D to 3D imaging. For 3D imaging in real-time, using the row-column (RC) addressing scheme, the number of elements scale linearly with the electrical interconnections. In contrast to the fully populated matrix array with a squared dependency, rendering it infeasible for large arrays. The row-column architecture was chosen for this project to facilitate 3D ultrasound, super-resolution, imaging in real-time. In order to achieve super-resolution many elements are required with a high yield, furthermore, a wide aperture or a large probe surface area is advantageous for maintaining a good focus throughout the human body.
The primary goal of this Ph.D. project is to develop large-scale 2D row-columnaddressed (RC) arrays of capacitive micromachined ultrasonic transducers (CMUTs) for 3D super-resolution ultrasound imaging in real-time of the erythrocytes to ultimately detect cancer and diabetes earlier. The CMUT platform was chosen for the development of row-column arrays due to its design flexibility offered by microfabrication technology, low self-heating, and wide bandwidth, all of which are advantageous for advanced imaging applications. This aim was accomplished through the successful design and fabrication of two large-scale chips resulting in an RC190+190 array and an RC512+512 array. The 2D CMUT arrays were achieved utilizing the anodic bonding fabrication technique with metal bottom electrodes to ensure a uniform ressure output even for large-scale arrays combined with an insulating substrate minimizing cross-talk between the elements. The process optimization performed in this Ph.D. study has resulted in the achievement of high-quality and void-free chip area of 7 cm × 7 cm, enabling the realization of future RC1024+1024 chips. However, the electrical yield is limited by vertical short circuits between the top and bottom electrodes due to the formation of gold defects, and future work should focus on eliminating the reaction forming the defects. One of the RC190+190 arrays was integrated into a modular prototype ultrasound probe designed for rapid prototyping within this research group. The electrical and preliminary acoustical performance of the RC190+190 CMUT was characterized. Further development is needed for optimal performance to be achieved. Additionally, the introduction, exploration and fabrication of a novel process flow of quartz fusion titanium disilicide CMUT was demonstrated with the potential to realize large-scale 2D CMUT arrays. Furthermore, specialized 3D-printed hydrogel phantoms were fabricated to validate the resolution and assess the performance of the advanced imagining algorithms. These phantoms, which are not commercially available, play a crucial role in achieving the objectives of the SURE project, namely, enabling 3D super-resolution ultrasound through large-scale RC CMUT probes and advanced imaging algorithms.