3D Ultrasound Accurately Measures Blood Flow

By Will Boggs MD

July 13, 2020

NEW YORK (Reuters Health) - Three-dimensional ultrasound can be useful for measuring constant and pulsatile volumetric blood-flow rates, researchers report.

"For a long time it was technically impossible or impractical to measure blood flow, that is, volumetric flow in milliliters per second, not velocity in centimeters per second. Ultrasound may bring a change to this," Dr. Oliver D. Kripfgans of Michigan Medicine, University of Michigan, in Ann Arbor, told Reuters Health by email.

Current blood-flow imaging methods are predominately qualitative. True sonographic measurements are rarely used clinically because of reliability issues.

Dr. Kripfgans and colleagues used a calibrated flow phantom to assess the accuracy of 3D color flow ultrasound (US) imaging for measuring quantitative volumetric blood flow under relevant flow rates, measurement depths, different users, and alterations in the flow past a stenosis.

They tested three clinical scanner models at three separate clinical sites.

The average absolute biases for constant flow were 6.3%, 8.5% and 16.6%, respectively, for the three systems, with coefficients of variation of 10.8%, 7.3%, and 6.7%, respectively, the researchers report in Radiology.

For pulsatile flow, the average biases were 18.5%, 9.0%, and 6.2%, respectively, with coefficients of variation of 4.3%, 6.7%, and 5.3%.

"The most interesting was to see that for one ultrasound scanner there was less than 5% variation between testing sites," Dr. Kripfgans said. "The sites used the same make and brand ultrasound scanner but physically different machines. This demonstrates the stability of the underlying concept and measurement."

Across flow rates ranging from 1 to 12 mL/sec, one system underestimated systematically for both constant and pulsatile flow, while another system overestimated for pulsatile flow conditions. Two of the three systems tracked within 10%.

Over depths ranging from 2.5 to 7 cm, coefficients of variation were mostly between 10% and 15%, except for outliers.

Stenosis can induce turbulence and other types of flow aberrations, but poststenotic constant and pulsatile flow biases were as low as 1.8% and 3.1%, respectively.

"Blood-flow estimation has a multitude of clinical applications, including more accurate blood flow measurement in hemodialysis arteriovenous fistulas and grafts, evaluating stenosis and function in transjugular intrahepatic portosystemic shunts, and other vascular systems where blood-flow quantification may be important to assess disease and response to treatment," the authors note.

They caution that "our study should be seen as preliminary work, which requires further investigations to address results that showed large bias or large coefficient of variation and to define acceptable flow estimation limits for bias and coefficient of variation for specific clinical applications because these results are only in a phantom."

"The results of the study by Kripfgans and colleagues will need in vivo verification," agrees Dr. Flemming Forsberg of Thomas Jefferson University, in Philadelphia, in a linked editorial.

"Important questions remain regarding the optimal real-time implementation across scanning platforms and how to accurately measure the true absolute blood flow and the sensitivity to slow flow (i.e., is there a need for contrast-enhanced US imaging?)," he said. "These questions need answers before the full potential of real-time in vivo 3D blood flow estimation can be achieved. It will also be crucial to compare real-time 3D volumetric US to other imaging modalities capable of estimating blood flow, such as dynamic MRI and CT angiographic techniques."

"Once this technique is incorporated onto clinical US scanners, and if these results can be reproduced in future clinical trials, quantitative 3D blood flow estimates will be one step closer to clinical practice, with the potential to markedly improve patient outcomes in many diseases," Dr. Forsberg concluded.

SOURCE: https://bit.ly/3eNU5ZX and https://bit.ly/3dKJvlq Radiology, online June 30, 2020.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....