Gespeichert in:
| Hauptverfasser: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2508.04051 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915430385319936 |
|---|---|
| author | Luo, Chen Jin, Qiyu Xie, Taofeng Wang, Xuemei Wang, Huayu Liu, Congcong Tang, Liming Chen, Guoqing Cui, Zhuo-Xu Liang, Dong |
| author_facet | Luo, Chen Jin, Qiyu Xie, Taofeng Wang, Xuemei Wang, Huayu Liu, Congcong Tang, Liming Chen, Guoqing Cui, Zhuo-Xu Liang, Dong |
| contents | Interpolating missing data in k-space is essential for accelerating imaging. However, existing methods, including convolutional neural network-based deep learning, primarily exploit local predictability while overlooking the inherent global dependencies in k-space. Recently, Transformers have demonstrated remarkable success in natural language processing and image analysis due to their ability to capture long-range dependencies. This inspires the use of Transformers for k-space interpolation to better exploit its global structure. However, their lack of interpretability raises concerns regarding the reliability of interpolated data. To address this limitation, we propose GPI-WT, a white-box Transformer framework based on Globally Predictable Interpolation (GPI) for k-space. Specifically, we formulate GPI from the perspective of annihilation as a novel k-space structured low-rank (SLR) model. The global annihilation filters in the SLR model are treated as learnable parameters, and the subgradients of the SLR model naturally induce a learnable attention mechanism. By unfolding the subgradient-based optimization algorithm of SLR into a cascaded network, we construct the first white-box Transformer specifically designed for accelerated MRI. Experimental results demonstrate that the proposed method significantly outperforms state-of-the-art approaches in k-space interpolation accuracy while providing superior interpretability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_04051 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Globally Predictable k-Space Interpolation: A White-box Transformer Approach Luo, Chen Jin, Qiyu Xie, Taofeng Wang, Xuemei Wang, Huayu Liu, Congcong Tang, Liming Chen, Guoqing Cui, Zhuo-Xu Liang, Dong Computer Vision and Pattern Recognition Optimization and Control Interpolating missing data in k-space is essential for accelerating imaging. However, existing methods, including convolutional neural network-based deep learning, primarily exploit local predictability while overlooking the inherent global dependencies in k-space. Recently, Transformers have demonstrated remarkable success in natural language processing and image analysis due to their ability to capture long-range dependencies. This inspires the use of Transformers for k-space interpolation to better exploit its global structure. However, their lack of interpretability raises concerns regarding the reliability of interpolated data. To address this limitation, we propose GPI-WT, a white-box Transformer framework based on Globally Predictable Interpolation (GPI) for k-space. Specifically, we formulate GPI from the perspective of annihilation as a novel k-space structured low-rank (SLR) model. The global annihilation filters in the SLR model are treated as learnable parameters, and the subgradients of the SLR model naturally induce a learnable attention mechanism. By unfolding the subgradient-based optimization algorithm of SLR into a cascaded network, we construct the first white-box Transformer specifically designed for accelerated MRI. Experimental results demonstrate that the proposed method significantly outperforms state-of-the-art approaches in k-space interpolation accuracy while providing superior interpretability. |
| title | Towards Globally Predictable k-Space Interpolation: A White-box Transformer Approach |
| topic | Computer Vision and Pattern Recognition Optimization and Control |
| url | https://arxiv.org/abs/2508.04051 |