现在你拥有我的健康注意力：用于脑部MRI修复的U-DiT

中文摘要 AI 辅助

ASNR-MICCAI BraTS局部合成（修复）任务要求在T1加权MRI的掩蔽区域内完成解剖学合理的健康脑组织，为下游分析提供无肿瘤的解剖学参考。由于该任务由失真指标（SSIM、PSNR、MSE）评分，我们构建了确定性回归模型，并专注于为其提供适配修复的归纳偏置。我们的网络遵循U-DiT原则，在降采样的令牌网格上执行自注意力：一个体积型编码器-解码器通过带有三维旋转位置嵌入的降采样全局自注意力块导入长程上下文，同时卷积和跳跃连接保留高频细节。两项思路推动了我们的结果：其一，我们约束注意力，使被遮挡（“空白”）令牌仅关注同一体积内已知的健康令牌，并学习每个查询对侧同源区域的偏置，迫使补全从观测到的解剖结构推断，而非从其他未知区域推断；其二，我们添加对侧对称输入，将镜像健康半球作为患者特定先验，由于大脑大致双侧对称且病变通常为单侧，该先验提升了匹配结构相似性下的失真指标。在官方BraTS-2026验证排行榜上，我们的提交在219例病例中达到健康区域平均SSIM为0.864、PSNR为24.7 dB、MSE为4.6×10⁻³。我们进一步分析了失真最优回归固有的残差平滑性，并讨论其对解剖学真实性的影响。

英文摘要

The ASNR-MICCAI BraTS Local Synthesis (Inpainting) task asks for the anatomically plausible completion of healthy brain tissue within a masked region of a T1-weighted MRI, providing a tumor-free anatomical reference for downstream analysis. As the task is scored by distortion metrics (SSIM, PSNR, MSE), we build a deterministic regression model and focus on giving it inductive biases tailored to inpainting. Our network follows the U-DiT principle of performing self-attention on a downsampled token grid: a volumetric encoder-decoder imports long-range context through a downsampled global self-attention block with three-dimensional rotary position embeddings, while convolutions and skip connections preserve high-frequency detail. Two ideas drive our results. First, we constrain the attention so that occluded ("void") tokens attend only to known-healthy tokens of the same volume, with a learned bias toward each query's contralateral homologue, forcing the completion to be inferred from observed anatomy rather than from other unknown regions. Second, we add a contralateral-symmetry input that supplies the mirrored healthy hemisphere as a patient-specific prior; since the brain is approximately bilaterally symmetric and lesions are typically unilateral, this prior improves the distortion metrics at matched structural similarity. On the official BraTS-2026 validation leaderboard our submission reaches a mean healthy-region SSIM of $0.864$, PSNR of $24.7$\,dB and MSE of $4.6{\times}10^{-3}$ over $219$ cases. We further analyse the residual smoothness inherent to distortion-optimal regression and discuss its implications for anatomical realism.

发表机构

• Politecnico di Bari（巴里理工大学）

机构由 AI 辅助整理，请以论文原文为准。