Keywords: Analysis/Processing, Spectroscopy, MRSI, Metabolite Maps
Motivation: Metabolite images from Magnetic Resonance Spectroscopic Imaging (MRSI) suffer from lower quality and reduced detail due to larger voxel sizes compared to anatomical MRI.
Goal(s): To improve the visual quality of MRSI by using a deep learning-based super-resolution approach to enhance spatial resolution.
Approach: Synthetic metabolic maps were generated using anatomical images from 350 patients. Our CNN-transformer model was trained on 70% of the dataset and tested on the remaining 30%, with performance compared to spline and nearest-neighbor methods.
Results: Our model significantly upscaled MRSI resolution to 128×128, achieving significantly higher PSNR, SSIM, and LPIPS scores than spline and nearest-neighbor (p<.01).
Impact: This improved SR approach significantly enhances metabolite map quality, offering clinicians a valuable tool for detailed neurological assessment.
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