Keywords: Analysis/Processing, Machine Learning/Artificial Intelligence, super-resolution
Motivation: High-resolution 3D MR imaging provides superior diagnostic quality compared to 2D scans, but its prolonged acquisition time limits its clinical adoption for all contrast weightings.
Goal(s): Develop a method that enhances the resolution of 2D FLAIR and T2 MR scans to similar levels as in 3D acquisitions.
Approach: Leveraging auxiliary high-resolution 3D images, we developed a model with a pre-trained masked autoencoder and residual convolutional decoder to enhance resolution for 2D MR scans.
Results: The developed method produces 3D images from 2D scans with similar quality comparable to 3D acquisitions, which can be 5- to 12-fold faster, enabling efficient multi-contrast MRI acceleration.
Impact: The developed approach allows multi-contrast brain low-resolution 2D scans with an auxiliary high-resolution 3D reference scan to produce multi-contrast high-resolution 3D images. Clinical evaluation on synthesized 3D brain images and extension to other applications may be worth further investigation.
How to access this content:
For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.
After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.
After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.
Keywords