Towards an Operator Independent Blood Flow Volume Quantification Using 3D Ultrasound

• Published in: 2020 IEEE International Ultrasonics Symposium (IUS) pp. 1 - 4
• Main Authors: Li, Sibo, Shi, William, Pinter, Stephen Z., Rubin, Jonathan M., Kripfgans, Oliver D., Fowlkes, J. Brian, Leichner, Ronald D., Jago, James R., Sethuraman, Shriram
• Format: Conference Proceeding
• Language: English
• Published: IEEE 07.09.2020
• Subjects: Color flow; Gaussian surface integral; Image color analysis; Phantoms; Reliability; Three-dimensional displays; Ultrasonic imaging; Ultrasonic variables measurement; Ultrasound; Volume flow; Volume measurement; xMATRIX transducer
• ISSN: 1948-5727
• DOI: 10.1109/IUS46767.2020.9251692

The measurement of blood volume flow is critical in many clinical applications. The current state-of-the-art for volume flow quantification uses pulsed-wave ultrasound, using a method that is neither accurate nor reliable. In this work, we implemented and demonstrated the feasibility of a 3D-based volume flow method on a Philips 3D xMATRIX ultrasound probe (XL14-3). To improve the workflow, an optimal scan sequence termed "iSTIC" was adopted for 4D color data acquisition. The acquisition time to obtain >20 volumes with high spatio-temporal resolution was controlled to within 15 seconds, which makes it feasible for clinical practice. This framework and repeatable measurements for volume flow estimation were tested on a customized flow phantom. Our experimental results showed comparable measurements between the 3D method and flow meter reference on the flow phantom. Under both constant and pulsatile flow conditions, we tested the performance of 3D volume flow by varying the color gain and pump flow rate. In constant flow, for the two aforementioned parameter changes, the measurements showed average biases of −2.7% and 2.9%, respectively. The corresponding coefficient of variation were 3.5% and 2.9%. Under pulsatile flow, the corresponding average biases were 3.3% and −3.6%, with coefficients of variation of 3.2% and 5.0%, respectively. These results indicate the potential of the 3D implementation to be a quick, reliable and accurate volume flow estimation technique and ready for the next step of clinical investigations.