Keywords: AI/ML Image Reconstruction, AI/ML Image Reconstruction
Motivation: High b-value data is crucial for Soma and Neurite Density Imaging (SANDI). However, the acquisition of high b-value images is highly dependent on the gradient performance of the equipment, making it not applicable to most clinical MRI systems.
Goal(s): This study presented a deep learning solution to synthesize high b-value images based on multi-phase initial low b-value images.
Approach: We proposed a multi-task learning image synthesis framework that combined generative adversarial loss and SANDI-sensitive loss.
Results: The proposed model achieved high-performance high b-value image synthesis and provided benefits for downstream SANDI map prediction.
Impact: This study explores the feasibility of replacing real high b-value images with synthesized images generated by deep learning models, which holds promise for transfer to other MRI systems with lower gradient performance, thereby expanding the application scope of SANDI.
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