Ultrasound beamforming with compressed data

Abstract

With the rapid advancement in electronics technologies, software-based beamformers for ultrasound array imaging have become feasible, and thus allowing rapid development of a high performance and potentially low cost system. However, to realize a fully software based system, one challenge is data transfer from the analog front end to the software backend, whose data rate can be up to a few GB/sec. It is therefore the purpose of this study to investigate the use of data compression for reducing data transfer requirements and the associated tradeoff with beamforming quality. JPEG (Joint Photographic Experts Group) and JPEG2000 compression were adopted. The acoustic data were acquired with a 128-channel array transducer at a center frequency of 3.5MHz. On each transmit event, the receive channel data are separated into 8×8 blocks, followed by the application of data compression. In one scheme, the compression is done on raw RF data. In another scheme, baseband data are used and only amplitude of the data is compressed. The maximum compression ratio of RF data compression is 15 with JPEG and 20 with JPEG2000 when the average error is less than 5 dB. With baseband data compression, on the other hand, although the maximum compression ratio is less than 8, the image quality is better than RF data compression. The average error in this case is less than 1.5 dB when the compression ratio is less than 7.