BEAM3R: Beam's-eye-view architecture with Mamba-3 for implicit dose reconstruction

📄 arXiv: 2609.04747v1 📥 PDF

作者: Chen Cheng, Michael Ferraro, James Grover, David E J Waddington, Emily Hewson

分类: physics.med-ph, cs.CV

发布日期: 2026-09-04

备注: 14 pages, 5 figures


💡 一句话要点

提出BEAM3R以解决剂量重建中的高效性与准确性问题

🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture)

关键词: 剂量重建 束眼视图 Mamba-3 深度学习 放射治疗 CT图像处理 物理基础建模

📋 核心要点

  1. 现有剂量计算方法在准确性和计算效率上存在不足,尤其是在处理复杂的束流传输时。
  2. BEAM3R通过束眼视图架构与Mamba-3核心的结合,优化了剂量重建过程,减少了计算复杂度。
  3. 在DoseRAD2026测试集上,CT到光子和质子模型的局部伽马通过率高达96.8%和96.0%,显示出显著的性能提升。

📝 摘要(中文)

为实现DoseRAD2026挑战中的准确快速的光子控制点和质子束剂量计算,本文提出了BEAM3R,一个基于束眼视图(BEV)的剂量估计框架。其核心创新在于结合Mamba-3状态空间深度序列核心与基于物理的传输条件,能够在不使用昂贵的3D卷积的情况下建模长程深度传输。BEAM3R采用2D CNN编码器-解码器架构处理光子和质子剂量任务,处理每平面BEV切片。实验结果表明,该框架在DoseRAD2026测试集上表现出色,CT到光子和质子模型的局部伽马通过率分别为96.8%和96.0%。

🔬 方法详解

问题定义:本文旨在解决在剂量重建中面临的计算效率和准确性问题,现有方法在处理长程深度传输时往往依赖昂贵的3D卷积,导致计算时间过长。

核心思路:BEAM3R框架通过束眼视图(BEV)与Mamba-3深度序列核心的结合,利用物理基础的传输条件来建模深度传输,从而提高计算效率和准确性。

技术框架:该框架采用2D CNN编码器-解码器架构,处理光子和质子剂量任务,使用每平面BEV切片进行数据处理。质子束流的输入参数包括水当量厚度和剩余范围,以编码布拉格峰位置。

关键创新:最重要的创新在于引入了Mamba-3核心与物理传输条件的结合,避免了3D卷积的高计算成本,同时实现了高空间分辨率的剂量重建。

关键设计:在设计中,采用了双向Mamba-3核心捕捉下游材料的剂量贡献,质子模型则使用前向核心与学习的能量前缀标记,结合布拉格峰细化模块,确保了模型的准确性。

🖼️ 关键图片

fig_0
img_1
img_2

📊 实验亮点

在DoseRAD2026测试集上,CT到光子和质子模型的局部伽马通过率分别达到了96.8%和96.0%,而使用合成CT时,伽马通过率下降至89.7%和75.4%。这些结果表明,BEAM3R在剂量重建任务中具有显著的性能优势。

🎯 应用场景

BEAM3R框架在放射治疗领域具有广泛的应用潜力,能够为临床提供快速而准确的剂量计算,优化治疗计划,提升患者的治疗效果。未来,该技术有望与其他成像技术结合,进一步提高剂量重建的精度和效率。

📄 摘要(原文)

To enable accurate and rapid photon control point and proton beamlet dose calculation in the DoseRAD2026 challenge, we present BEAM3R, a dose estimation framework operating in beam's-eye-view (BEV). Our core innovation combines a Mamba-3 state-space depth-sequence core with physics-based transport conditioning to model long-range depth transport without expensive 3D convolutions. BEAM3R shares a 2D CNN encoder-decoder architecture for photon and proton dose tasks, processing per-plane BEV slices. Proton beamlets are conditioned on water equivalent thickness and remaining range, encoding the parameters determining Bragg peak position. Photon models use a bidirectional Mamba-3 core to capture dose contributions from materials downstream of the calculation point, while the proton model uses a forward core with learned energy-prefix tokens and a Bragg-peak refinement module. To reduce interpolation artifacts and support high spatial resolution, we introduce axial grid alignment of BEV lattices with CT slices and an implicit super-resolution representation via sub-pixel phase packing, evaluated by a differentiable Triton-accelerated resampler that reconstructs packed cubic B-spline coefficients directly in CT space. For MRI-based tasks, synthetic CTs (sCT) are generated by a patch-based conditional GAN with a SwinUNETR backbone. On the preliminary DoseRAD2026 test set, CT-to-photon and CT-to-proton models achieved 1%/1 mm local gamma pass rates of 96.8% and 96.0%, with stratified plan-level MAEs of 0.0041 and 0.0079. Substituting sCT reduced gamma pass rates to 89.7% for photon and 75.4% proton plan level doses, with stratified plan-level MAEs of 0.0093 and 0.0336. Standardised runtimes were 23.4 s and 18.4 s for CT-to-photon and CT-to-proton prediction, increasing to 39.7 s and 42.8 s for the corresponding MRI-based pipelines.