Send to

Choose Destination
Z Med Phys. 2014 Sep;24(3):190-200. doi: 10.1016/j.zemedi.2013.10.005. Epub 2013 Nov 13.

Accelerated radial Fourier-velocity encoding using compressed sensing.

Author information

Institute of Radiology, University of Würzburg, Germany. Electronic address:
Institute of Radiology, University of Würzburg, Germany; Comprehensive Heart Failure Center (CHFC), University of Würzburg, Germany.
Institute of Radiology, University of Würzburg, Germany.



Phase Contrast Magnetic Resonance Imaging (MRI) is a tool for non-invasive determination of flow velocities inside blood vessels. Because Phase Contrast MRI only measures a single mean velocity per voxel, it is only applicable to vessels significantly larger than the voxel size. In contrast, Fourier Velocity Encoding measures the entire velocity distribution inside a voxel, but requires a much longer acquisition time. For accurate diagnosis of stenosis in vessels on the scale of spatial resolution, it is important to know the velocity distribution of a voxel. Our aim was to determine velocity distributions with accelerated Fourier Velocity Encoding in an acquisition time required for a conventional Phase Contrast image.


We imaged the femoral artery of healthy volunteers with ECG-triggered, radial CINE acquisition. Data acquisition was accelerated by undersampling, while missing data were reconstructed by Compressed Sensing. Velocity spectra of the vessel were evaluated by high resolution Phase Contrast images and compared to spectra from fully sampled and undersampled Fourier Velocity Encoding. By means of undersampling, it was possible to reduce the scan time for Fourier Velocity Encoding to the duration required for a conventional Phase Contrast image.


Acquisition time for a fully sampled data set with 12 different Velocity Encodings was 40 min. By applying a 12.6-fold retrospective undersampling, a data set was generated equal to 3:10 min acquisition time, which is similar to a conventional Phase Contrast measurement. Velocity spectra from fully sampled and undersampled Fourier Velocity Encoded images are in good agreement and show the same maximum velocities as compared to velocity maps from Phase Contrast measurements.


Compressed Sensing proved to reliably reconstruct Fourier Velocity Encoded data. Our results indicate that Fourier Velocity Encoding allows an accurate determination of the velocity distribution in vessels in the order of the voxel size. Thus, compared to normal Phase Contrast measurements delivering only mean velocities, no additional scan time is necessary to retrieve meaningful velocity spectra in small vessels.


Beschleunigte Aufnahme; Compressed Sensing; Geschwindigkeitsmessung; Herz-Kreislauf; Radial MRI; Radiale MR-Bildgebung; accelerated acquisition; cardio-vascular; velocity mapping

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center