Communications of the ACM

Home/News/AI Accelerates Blood Flow MRI/Full Text

ACM TechNews

AI Accelerates Blood Flow MRI

By ETH Zurich
April 30, 2020
Comments

View as: Print Mobile App Share:

The new method makes it possible to obtain precise MRI images of blood flow in a sixth the time it takes now. — ETH Zurich and the University of Zurich researchers have developed an artificial intelligence program that could greatly accelerate dynamic magnetic resonance imaging of blood flow in the human body.

Credit: CMR Zurich

Researchers at ETH Zurich and the University of Zurich in Switzerland have developed an artificial intelligence program that could greatly accelerate dynamic magnetic resonance imaging (dMRI) of blood flow.

Currently, the recording and subsequent processing of a four-dimensional (4D) flow MRI takes up to 30 minutes, while the new program, called FlowVN, could accomplish the same process within five minutes.

FlowVN is different from other deep learning programs because it combines training with prior knowledge of the measurement, which makes it extremely efficient.

The researchers trained the software using 11 MRI scans of healthy test subjects. This data was used to accurately reproduce pathological blood flow in a patient's aorta on an ordinary computer within just 21 seconds.

Said ETH Zurich's Sebastian Kozerke, "We hope that FlowVN will drive forward the use of 4D flow MRI in clinical diagnostics."

From ETH Zurich
View Full Article

No entries found