You are here : Home > News > Understanding the Origin of Artifacts in Ultra-High-Field MRI to Better Eliminate Them

Scientific result | Article | Brain | MRI

Understanding the Origin of Artifacts in Ultra-High-Field MRI to Better Eliminate Them


Ultra-high-field MRI enables unprecedented precision in brain imaging, but it is not immune to artifacts. Among the many potential sources of interference, a team at NeuroSpin has focused on a specific phenomenon: artifacts related to intra-voxel dephasing, which occur when using accelerated acquisition methods like GRAPPA. Their work sheds light on the mechanisms behind these distortions and proposes solutions to mitigate them.

Published on 16 February 2026
​

Ultra-high-field MRI scanners, such as the 11.7 T Iseult MRI at NeuroSpin, provide images with exceptional resolution. However, at these field strengths, various artifacts can degrade data quality. Researchers from METRIC team (Mixed unit BAOBAB/NeuroSpin) have investigated a particular type of distortion: wave-like artifacts caused by intra-voxel dephasing in regions where the magnetic field is inhomogeneous.

Why are these artifacts​​ problematic?

To accelerate acquisitions, scientists use algorithmic and sampling techniques like GRAPPA, which reconstruct images from partially acquired data. The calibration lines (ACS), essential for this reconstruction, are often low-resolution and more sensitive to local variations in the magnetic field. In areas where these variations are strong (such as near the sinuses or the neck), the signal dephases within the voxels themselves, creating interference that propagates and blurs the final image.

How do researchers pro​​pose to mitigate them?

The team demonstrates that:

  • External ACS (with a short echo time) significantly reduce artifacts, as they better preserve the signal in disturbed areas.
  • Increasing the number of ACS lines improves image quality, though at the cost of a slightly longer acquisition time.
  • Optimizing reconstruction parameters (such as kernel size) helps limit artifacts, although it does not eliminate them entirely.
  • Selective excitation (to avoid problematic areas) is an effective solution but requires case-specific adjustments.

A challenge for research​ and medicine

These artifacts can distort results, particularly in studies on neurodegenerative diseases or brain mapping. By understanding their origin, researchers enhance the reliability of images—a major asset for research and clinical applications.​

Contact Frédéric-Joliot Institute for Life sciences:

Nicolsa Boulant (n​​​icolas.boulant@cea.fr)

European fundings
This work was carried out as part of the FET OPEN project "AROMA", coordinated by Nicolas Boulant.


PIA3 Fundings
The project received support through PRESENCE, funded by the "Structuring Equipment for Research: EquipEx+" initiative​.
​​​This text was translated with the assistance of Mistral AI.

Top page