用于自回归MRI重建的下一加速倍率预测

摘要

MRI重建本质上是一个不适定逆问题，因为不完整的测量数据对应着多种合理的解。这种模糊性在高加速倍数下尤为严重，此时像素域连续预测器倾向于在可行重建结果中求平均，从而抑制了高频解剖结构。针对这一局限，我们将重建过程转移至离散多尺度潜空间，并将其建模为下一加速尺度的自回归预测问题。借助视觉自回归建模中已被验证有效的离散先验，我们的方法将解空间约束为紧凑的码本标记序列，即便在极度稀疏的测量条件下也能生成锐利重建。这种离散自回归形式也自然契合现代大型语言模型的后训练技术。基于这一观察，我们提出了面向视觉自回归建模的在策略特权信息蒸馏方法：教师模型仅利用推理阶段不可用的特权上下文（在本工作中指全采样采集数据）进行训练，并监督在其自身生成序列上训练的学生模型，从而获得一致的重建性能提升。通过在fastMRI基准上的大量实验，我们证明该方法在极低欠采样率下的多种采样模式中均能实现更优的重建性能。项目网站见：https://yilmazkorkmaz1.github.io/discrete-mri-reconstruction-opd/{here}。

English

MRI reconstruction is an inherently ill-posed inverse problem, since incomplete measurements admit many plausible solutions. This ambiguity becomes more severe under high acceleration, where pixel-domain continuous predictors tend to average over feasible reconstructions and suppress high-frequency anatomy. We address this limitation by moving reconstruction to discrete multi-scale latent space and posing it as autoregressive next-acceleration-scale prediction. Leveraging discrete priors proven effective in visual autoregressive modeling, our method restricts the solution to compact sequences of codebook tokens, enabling sharp reconstructions even from extremely sparse measurements. This discrete autoregressive formulation also aligns naturally with modern large language model post-training techniques. Building on this observation, we introduce on-policy privileged information distillation for visual autoregressive modeling, where a teacher is provided training only privileged context that is unavailable at inference, in our case fully sampled acquisitions, and supervises a student trained on its own rollouts, leading to consistent reconstruction gains. Through extensive experiments on the fastMRI benchmark, we show that our approach delivers improved reconstruction performance across diverse sampling patterns under extreme undersampling. Project website is https://yilmazkorkmaz1.github.io/discrete-mri-reconstruction-opd/{here}.