Micromachined integrated 2D transducers for super resolution ultrasound imaging

Medical ultrasound imaging is a widely used real-time technique for non-invasive diagnostics. Conventional ultrasonics 1D linear probes produce 2D images of the human body. Organs and tissues are, however, not confined to a 2D plane in the body. Vital information out of plane may therefore be lost due to the three dimensional nature of the organ and any movement of the probe. A major drive in the field of ultrasound is therefore to progress towards 3D ultrasound imaging using 2D arrays. To achieve a good focus and a high resolution, a large number of elements are needed, which translates to probes with large surface areas. In fully addressed 2D matrix arrays, this increases the complexity of the underlying technology. The number of interconnections needed in probe will also scale with, N², which quickly makes the process infeasible for large arrays with a high channel or element count, N.
Recently, a different type of technology using a row-addressing element scheme has been introduced. This reduced the complexity of the arrays by addressing only the N row and N column elements, and the required number of interconnects is only 2N.
The main goal of this project has to develop large scale 2D 190+190 row-column-addressed (RCA) capacitive micromachined ultrasonic transducer (CMUT) arrays for the use in three dimensional real-time volumetric imaging. The CMUT platform was chosen as offers a high degree of flexibility, low self-heating and a large bandwidth beneficial for imaging applications. This has been achieved through the successful fabrication of two chip designs. These were based on two different fabrication techniques, and investigated to produce stable and reliable transducers. The techniques are using the local oxidation of silicon (LOCOS) based process combining fusion and anodic bonding with highly doped silicon as bottom electrodes, and using a purely anodic bonding process with metal bottom electrodes. They each have their advantages as CMUT platforms for building arrays with a uniform pressure output for all elements.
These chips were integrated in custom made hand-held transducer probes. One of these probes being a modular prototype probe for rapid prototyping developed in this research group. The electrical and acoustical performance of the two transducers were evaluated, and the 3D imaging capabilities for one of them was demonstrated. The imaging depth was found to be 3.6 cm. In conclusion, results show that the row–column technology is a realistic alternative to matrix probes for volumetric imaging. Furthermore, it is shown that the two processing techniques can be used as viable platforms for producing stable transducers. Further development is needed for achieving optimal performance.