扩散磁共振 Transformer 与扩散空间旋转位置嵌入（D-RoPE）
Diffusion MRI Transformer with a Diffusion Space Rotary Positional Embedding (D-RoPE)
D-RoPE 把扩散梯度方向与强度编码进旋转位置嵌入，使 Transformer 能在任意数量扩散方向上共享权重；做多通道/多视角输入（如多光照、多视角图像）时，可借鉴这种把采集参数写进位置编码以摆脱固定通道数的做法。
原文依据 we introduce a diffusion space rotatory positional embedding (D-RoPE) plugged into our dMRI transformer
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论扩散磁共振成像（dMRI）在研究大脑微结构变化方面起着关键作用，因而被广泛应用于临床实践；然而，从 dMRI 中学习通用表征的进展一直有限。一个关键挑战在于，现有的深度学习方法难以充分捕捉扩散信号的独特性质。大脑 dMRI 通常由若干个体数据组成，每个体数据具有不同的衰减特性，这些特性取决于扩散敏感梯度的方向与强度。因此，需要对 dMRI 中的空间依赖、扩散加权依赖与方向依赖进行联合建模。此外，不同的采集协议（例如方向数量不同）进一步限制了传统模型。为弥补这些不足，我们提出一种扩散空间旋转位置嵌入（D-RoPE），并将其接入我们的 dMRI Transformer，以同时捕捉扩散数据的空间结构与方向特性，从而在不同采集设置以及任意数量的扩散方向下获得稳健且可迁移的表征。经过自监督掩码自编码预训练后，在多个下游任务上的测试表明，所学表征与预训练模型在这些下游任务上可取得与多个基线相当或更优的性能（甚至优于一个完全训练的基线）；由我们预训练编码器微调得到的特征在轻度认知障碍分类上准确率提升 6%，在认知评分预测上相关系数提升 0.05。代码见：github.com/gustavochau/D-RoPE
英文原文摘要 1707 字符
Diffusion Magnetic Resonance Imaging (dMRI) plays a critical role in studying microstructural changes in the brain. It is, therefore, widely used in clinical practice; yet progress in learning general-purpose representations from dMRI has been limited. A key challenge is that existing deep learning approaches are not well-suited to capture the unique properties of diffusion signals. Brain dMRI is normally composed of several brain volumes, each with different attenuation characteristics dependent on the direction and strength of the diffusion-sensitized gradients. Thus, there is a need to jointly model spatial, diffusion-weighting, and directional dependencies in dMRI. Furthermore, varying acquisition protocols (e.g., differing numbers of directions) further limit traditional models. To address these gaps, we introduce a diffusion space rotatory positional embedding (D-RoPE) plugged into our dMRI transformer to capture both the spatial structure and directional characteristics of diffusion data, enabling robust and transferable representations across diverse acquisition settings and an arbitrary number of diffusion directions. After self-supervised masked autoencoding pretraining, tests on several downstream tasks show that the learned representations and the pretrained model can provide competitive or superior performance compared to several baselines in these downstream tasks (even compared to a fully trained baseline); the finetuned features from our pretrained encoder resulted in a 6% higher accuracy in classifying mild cognitive impairment and a 0.05 increase in the correlation coefficient when predicting cognitive scores. Code is available at: github.com/gustavochau/D-RoPE