Harmonizing 2D and 3D FLAIR MRIs in white matter hyperintensity quantification
Motivation: Harmonizing neural imaging datasets respectively acquired with 2D and 3D FLAIR MRI. Goal(s): Converting 2D FLAIRs to high-resolution 3D FLAIRs. Approach: We employed a ResUNet-based deep learning approach to learn the complex transformation from 2D to 3D FLAIR. Results: The converted 3D FLAIRs bear a high resemblance to the acquired 3D FLAIR in terms of image similarity measures and white matter hyperintensity segmentation. Impact: With this proposed approach, we can harmonize the 2D FLAIRs from the ADNI study with the 3D FLAIRs in the UK Biobank study.
Authors
- Andria L. Ford (ORCID: https://orcid.org/0000-0003-0605-8943)
- Jin‐Moo Lee (ORCID: https://orcid.org/0000-0002-3979-0906)
- Yasheng Chen (ORCID: https://orcid.org/0000-0002-1133-452X)
- Chunwei Ying (ORCID: https://orcid.org/0000-0003-1060-3289)
- Hongyu An (ORCID: https://orcid.org/0000-0001-6459-2269)
- Peter Kang (ORCID: https://orcid.org/0000-0002-9823-4440)
- Chia‐Ling Phuah (ORCID: https://orcid.org/0000-0002-6705-7621)
- Xing Dai (ORCID: https://orcid.org/0000-0001-6550-0880)
Publication Details
- Journal
- Proceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition
- Published
- 2024-11-26
- DOI
- https://doi.org/10.58530/2024/0748
- Primary Topic
- Advanced Neuroimaging Techniques and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00