Keywords: AI/ML Software, Diffusion Denoising
Motivation: Proprietary denoising methods using deep learning have been recently released for clinical use, but their performance against open source denoising methods in diffusion MRI (dMRI) is unclear.
Goal(s): To compare efficacy of GE HealthCare’s AIR-ReconDLTM (ARDL) and patch-based MPPCA and NORDIC denoising for reducing noise-induced variance and bias in high-resolution dMRI data.
Approach: We compared different denoising approaches, both in complex and magnitude domains, considering metrics reflecting noise variance and noise floor suppression.
Results: ARDL denoising, which operates in the complex domain. had the most noise-floor suppression. NORDIC complex denoising had the highest gain in SNR, with comparable, yet improved, signal dynamic range.
Impact: Our results suggest that denoising in the complex domain compared to magnitude domain has the potential to lead to larger denoising benefits than any differences induced by the employed denoising approach (e.g. deep learning vs patch-based).
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