Development of low cost, high channel count research platform for democratizing 3D volumetric ultrasound - SUMMARY ABSTRACT Three-dimensional (3D) imaging results in better diagnostics and clinical benefit. However, 3D ultrafast volumetric imaging using ultrasound has lagged other modalities such as MRI and X-ray CT. Ultrafast volumetric imaging enables functional imaging modes like volumetric color flow imaging, volumetric shear wave imaging, and volumetric functional ultrasound. The major factor hindering the development of 3D ultrafast ultrasound is the cost associated with hardware platforms that can control probes (2D arrays) with large channel counts, i.e., 1024 channels or more. To overcome these issues and lower costs, different solutions have been proposed that result in tradeoffs in image quality and speed of image generation, e.g., freehand movement of linear arrays, mechanical translation of linear arrays, row-column arrays. However, these solutions are sub-optimal. The optimal solution, in terms of performance, is a 2D array. A 2D array can provide ultrafast volumetric imaging, low side lobes, improved contrast, and a consistent point spread function in both lateral and elevation directions. The tradeoffs with a 2D array are that it requires a larger element count and channel count to operate the array, which is more expensive to manufacture, both for the array and the electronics to operate the array. The cost of a 2D array (32x32 elements) is $20-30k or more, which makes it an expensive array, but still affordable. Although multiplexing solutions are available with a tradeoff in volume rate, to fully operate the array with its maximum potential, each element should be controllable through electronics. Currently, to operate a 32x32 element array would require 1024 channels. One way that this has been accomplished is by linking several research imaging platforms together, such as 4x256 Verasonics Vantage systems. The cost of purchasing and combining four of these systems together is ~$1,000,000, which is out of reach for many ultrasound research laboratories and is cost prohibitive for use in clinical scenarios. These high costs have limited access to these devices and hindered progress in ultrafast, volumetric ultrasound imaging capabilities and innovations. Our proposed project is to develop a low-cost, novel, open-source ultrasound imaging research platform with 1024 individually addressable channels that emphasizes the production of fast 3D volumetric data acquisition and image formation, i.e., real time and ultrafast volumetric imaging. We estimate the cost of hardware to produce our 1024-channel ultrasound research platform will be <$70k. To accomplish this goal, four specific aims are proposed. In the first specific aim, we will design and validate our FPGA-based distributed volumetric beamformer and ultrasound over ethernet protocol. This will allow real time volumetric imaging and a low power budget. The second aim is to design, construct, and test a 1024-channel ultrasound scanning system capable of operating a 32x32 2D array at a cost of under <$70k. The third aim is to develop a user-friendly toolchain to allow user-defined customized imaging tasks. The final aim is to demonstrate in vivo real-time ultrafast volumetric imaging using a 2D matrix array and our novel system. The new system will democratize research in 3D volumetric imaging for ultrasound and facilitate innovations in this critical area.
