Imaging transformer for MRI denoising with SNR unit training: enabling generalization across field-strengths, imaging contrasts, and anatomy
Motivation: MR denoising using the CNN models often requires abundant high quality data for training. In many applications, such as higher acceleration and low field, high quality data is not available. This study overcome this limitation by developing a SNR unit based training scheme and a novel imaging transformer (imformer) architecture. Goal(s): To develop and validate a novel imformer model for MR denoising, enabling generalization across field-strengths, imaging contrasts, and anatomy. Approach: SNR unit training scheme and imaging transformer architectureResults: Imformer models outperformed CNNs and conventional transformer. The SNR training enables storng generalization. Impact: Recovery high-fidelity MR signal from very low SNR inputs; Enable 0.55T MRI model training.
Authors
- James Moon (ORCID: https://orcid.org/0000-0001-8071-1491)
- Iain Pierce (ORCID: https://orcid.org/0000-0001-8546-5023)
- Peter Kellman (ORCID: https://orcid.org/0000-0002-9875-6070)
- Sarah Hooper (ORCID: https://orcid.org/0000-0001-9366-2174)
- Azaan Rehman
- Thomas A. Treibel (ORCID: https://orcid.org/0000-0003-1560-7414)
- Rajiv Ramasawmy (ORCID: https://orcid.org/0000-0002-9821-9187)
- W. Patricia Bandettini (ORCID: https://orcid.org/0000-0003-2342-5637)
- Hui Xue (ORCID: https://orcid.org/0000-0002-4561-5530)
- Ahsan Javed (ORCID: https://orcid.org/0000-0003-1311-1247)
- Adrienne E. Campbell‐Washburn (ORCID: https://orcid.org/0000-0002-7169-5693)
- Rhodri Davies (ORCID: https://orcid.org/0000-0001-7630-7517)
- Yang Yang (ORCID: https://orcid.org/0009-0004-2290-805X)
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/10.58530/2024/0654
- Citations
- 1
- Primary Topic
- Advanced MRI Techniques and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.40