Comparison of echo-power estimation using linearized video data and raw data for Dynamic Contrast-Enhanced Ultrasound

To estimate perfusion using Dynamic Contrast-Enhanced Ultrasound (DCE-US), some manufacturers provide software that analyses DICOM Raw Data formatted files. In those files, most of the information cannot be correctly interpreted except by the manufacturer software. This approach when available is often recommended over use of more accessible DICOM JPEG files which encapsulate ultrasound system images as compressed JPEG images. However, the manufacturers' dedicated software is machine specific with limited functionalities which hinders multicentric studies using different ultrasound machines. The objective of this work is to develop techniques and criteria defining conditions under which compressed video images obtained with the Aplio 50 from Toshiba can be used to accurately estimate contrast concentration. Dedicated CHIQ software from Toshiba was compared to two approaches based on compressed images: commercially available VueBox™ software from Bracco and in-laboratory software PixPower. Proportionality between echo-power estimates and dose of contrast agents in suspension was validated in vitro for the three approaches. VueBox™ and PixPower echo-power estimations from DCE-US sequences acquired in an in vivo tumor model were in agreement with those obtained with the CHIQ reference. This work shows that, for the dynamic ranges considered in this study, DICOM JPEG files can be used to accurately estimate the echo power from video compressed images using linearizing equations based on rigorous calibration steps.