超声成像中高效的传感器利用与先进弹性图图像生成方法

Ultrasound is an important imaging technique mainly due to its non-intrusive nature. Due to high frame frequency in plane wave ultrasound imaging, a large number of raw data frames are being generated within a few seconds. In this technique, all transducer sensors are being used for imaging purposes. As a result, the amount of channel data acquired is large. Because of which, saving this data and transmitting it from ultrasonic system to processing computer is presenting a significant problem. To address this large data volume problem, we proposed the design of a novel transducer and implemented it. The proposed transducer is equipped with switching networks, which help in generating a minimized version of raw data in the front-end ultrasonic system, followed by reconstructing that missing channel data using Fourier and Inverse Fourier transforms inside the computer block. Next, in order to further reduce the amount of channel data capturing, sampling rate of that transducer was reduced. While the prior lowered the size of raw data by 50%, the latter reduced it by 75%. Note that the data transfer efficiency has also increased by same amounts, when compared to full data. These savings come with a slight increase in computer’s processing time, presented as software run time. Inside the computer block, beamformed, displacement and elastogram images are generated after Fourier based data reconstruction. Lastly, we focused on elastogram imaging and generated better quality elastographic images when compared to existing method, by means of quantization and colormap adjustment. While the soft and hard structures are clearly distinguishable in all our elastogram images, this is difficult in images generated using ground truth.