超越稀疏性：剪枝MRI重建中的权重位置与网络上下文
Beyond Sparsity: Weight Location and Network Context in Pruned MRI Reconstruction
浏览论文内容
中文总结 AI 辅助
该研究通过120个U-Net和视觉变换器模型，发现剪枝后MRI重建质量不仅取决于稀疏度，还受权重位置和网络上下文影响，提出架构特定描述符可辅助评估掩码质量。
中文摘要 AI 辅助
剪枝减少了磁共振成像（MRI）重建网络中的权重数量。然而，相同的稀疏度可以保留具有不同计算角色以及与训练网络不同兼容性的权重。我们在120个卷积U-Net和视觉变换器模型上研究了这些效应，使用在重训练前评估的受控编辑。在U-Net中，在相同删除数量下，保留高分辨率计算在所有测试的稀疏度水平上改善了重建效果。等操作对照揭示了第一卷积的额外敏感性，这仅凭操作数量无法解释。在变换器中，保留的预训练权重能量与稀疏度水平内的质量相关，然而当周围训练权重改变时，相同的掩码会反转其重建排名。在90%稀疏度下，匹配中间输出规模在很大程度上消除了高能量掩码的惩罚，同时保留了网络原始掩码的优势。因此，仅稀疏度并不能表征重建质量：架构特定的描述符具有信息性，但掩码质量仍可能取决于周围的训练网络。
英文摘要
Pruning reduces the number of weights in magnetic resonance imaging (MRI) reconstruction networks. Equal sparsity, however, can retain weights with different computational roles and different compatibility with the trained network. We study these effects across 120 convolutional U-Net and vision transformer models using controlled edits evaluated before retraining. In U-Net, preserving high-resolution computation improves reconstruction at equal deletion counts across all tested sparsity levels. Equal-operation controls reveal additional sensitivity of the first convolution, which operation count alone cannot explain. In transformers, retained pretrained weight energy correlates with quality within sparsity levels, yet the same masks reverse their reconstruction ranking when the surrounding trained weights change. At 90\% sparsity, matching intermediate output scales largely removes a high-energy-mask penalty while retaining an advantage for the network's original mask. Thus sparsity alone does not characterize reconstruction quality: architecture-specific descriptors are informative, but mask quality can still depend on the surrounding trained network.
发表机构
- Technion - Israel Institute of Technology（以色列理工学院）
机构由 AI 辅助整理，请以论文原文为准。
