Design of a multi-channel pre-beamform data acquisition system for an ultrasound research scanner

Abstract

Access to the pre-beamform data of each array channel on an ultrasound scanner is important to experimental investigations on advanced imaging research topics like adaptive beamforming and synthetic aperture imaging. Through such data access, we can obtain in-vitro or in-vivo insights on various imaging methods without resorting to hardware implementation. This paper reports the development of a pre-beamform data acquisition (DAQ) system that can collect data from 128 array elements in parallel. Our DAQ system is intended to interface with a Sonix-RP research scanner through a probe connector port. It comprises three major blocks: 1) a connector board that interfaces with the array probe and the scanner; 2) a main board that triggers data acquisition and controls data transfer to a computer; 3) four receiver boards that are each responsible for acquiring 32 channels of digitized raw data and storing them to the on-board DDR2 memory. The probe-connecting end of this system is interfaced with TX810 chips to facilitate switching between transmit and receive modes. When receiving data, the incoming analog signal is first passed into a low-noise amplifier. These signals are then sent into AD9272 chips on the four receiver boards to perform time gain compensation, anti-alias filtering, and 12-bit data sampling at an 80MHz rate. Subsequently, the digitized signal samples are de-serialized using a Virtex-5 FPGA and are stored into DDR2 memory. To facilitate data retrieval, we implemented a Virtex-5 FPGA on the main board to initiate data transfer, and if preferred, to perform onboard beamforming (programmable by user) prior to sending data to a computer through a USB 2.0 link. For our prototype, we used 16GB of DDR2 memory for data storage. With a frame rate of 35Hz, a 10cm depth-of-view and 128 transmit firings per frame, this DAQ system is capable of collecting pre-beamform data from all array channels for 2.5 seconds. We are currently completing the prototype development in collaboration with Ultrasonix.

Real-time GPU-based software beamformer designed for advanced imaging methods research

“…To evaluate the computational performance of our GPUbased beamformer prototype, a series of processing trials has been carried out. These trial runs are conducted on SAI and PWI test data acquired using a pre-beamform data acquisition system [15] that is connected to a reconfigured research scanner (Sonix-RP; Ultrasonix, Richmond, Canada). The RF sampling rate for the test datasets is 40 MHz, and each frame of pre-beamform data (i.e.…”

Blood flow image by multi-angle composite ultrasonic Doppler vector

“…Dunmire 等 [19] 则结合平面波传输提出了交叉波束的多普勒 方法, 通过发射两个对称的平面波束实现在波束交 叉区域的二维速度矢量成像. 为进一步提升成像帧 率, 子孔径传输的矢量多普勒逐渐兴起 [20,21] , 通过 设置特定子孔径偏转接收来估算不同方向上的速 度, 进而推导出整个流场信息, 且在短时间内获取 大量的数据. 另外, 基于多波束传输原理的矢量多 普勒方法在重建二维流场方面效果显著 [22,23]…”

FPGA-based reconfigurable processor for ultrafast interlaced ultrasound and photoacoustic imaging

“…), which is the driving need for a reconfigurable processor to switch between the two modalities in real-time using the same receiving frontend, and to manage the digital resources for optimizing the speed in each mode. Furthermore, the FPGA code allows real-time acquisition of pre-beamformed data for both modalities [18], this capability will provide researchers with an extended tool for forming their own images using extended imaging algorithms. Though the overall achieved co-registered frame rate is 1 Hz, this is primarily because of the use of the current DSP board, which has a slow link to the host PC.…”