Three dimensional flow and micro-flow imaging using Row-Column Arrays

Three-dimensional ultrasonic imaging and flow estimation offer a huge potential for clinicians, as it is a safe, reliable way of showing anatomical structures and the dynamic changes in blood flow patterns in real time and in full 3-D. Row column (RC) addressed transducer arrays have in recent years attracted a growing interest, as they greatly reduce the required channel count, allowing large aperture arrays to be used without requiring expensive and complex hardware. However, they also impose a limitation in terms of beam-steering, which has discouraged the use of them. This PhD seeks to investigate whether RC arrays can be employed for 3-D ultrasonic anatomical imaging, full 3-D vector flow estimation over time, here named tensor velocity imaging (TVI), and microvascular imaging (3-D super resolution imaging SRI), using the same synthetic aperture (SA) sequence.
Initially, a large-scale 192+192 channel RC array has been investigated for its B-mode capabilities, using simulations. The probe design utilized a novel zig-zag element pattern directly enabled by the use of capacitive micro-machined ultrasonic transducers (CMUTs). The design was verified as the image performance did not change compared to the standard straight design. Likewise, B-mode showed good contrast and resolution (dynamic range of 60 dB) while covering a large field-of-view, ideal for 3-D anatomical imaging.
Next, it has been investigated whether TVI can be achieved with a 62+62, 3 MHz, RC prototype array probe. A novel SA sequence that combines volumetric B-mode and TVI is presented. The sequence has been investigated through simulations and measurements. The study concludes that TVI is possible with simultaneous volumetric B-mode guidance at high volume rates. All three velocity components were estimated with a precision lower than 10%, at a peak velocity of 25 cm/s. The estimator was, however, sensitive to smaller beam-to-flow angles which decreased accuracy slightly.
Lastly, 3-D SRI has been investigated with the use of the same 62+62 RC array. A pulse inversion SA sequence is implemented to achieve full volumetric data. The method is tested using a validation platform enabled by a 3-D PEDGA printing process, where structures in 10-200 μm range can be printed. Localization precision of 23 μm was achieved, which is well below the diffraction limit resolution of the 3 MHz (500 μm).
The work concludes that a 62+62, 3 MHz RC array with the current SA sequence design can achieve full volumetric B-mode, TVI and 3-D SRI using only 124 channels. The larger 192+192 RC array is believed to provide new opportunities in 3-D imaging, although it remains to be investigated experimentally once it arrives from fabrication.