Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2609.04747

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

Chen Cheng·Michael Ferraro·James Grover·David E J Waddington·Emily Hewson
View original source ↗Read source document ↗Export metadata ↓

Abstract

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.

Explore related subjects

physics.med-phcs.CV

Keep this discovery

BibTeXRIS

Chen Cheng, Michael Ferraro, James Grover, David E J Waddington, Emily Hewson. 2026-09-04. BEAM3R: Beam's-eye-view architecture with Mamba-3 for implicit dose reconstruction. https://arxiv.org/abs/2609.04747

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

cs.CV subjectphysics.med-ph subject

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related papers

arXiv2026

MRpro: open framework for model-based, learned, and quantitative MR imaging

We preseent an open-source image reconstruction package built upon PyTorch, enabling modern deep-learning reconstructions. It uses open data formats for input and output (ISMRMRD, DICOM, NIfTI), allowing easy integration into existing pipelines and support for data from different devices. The framework comprises three main areas. First, it provides unified data structures for the consistent manipulation of MR datasets and their associated metadata (e.g., k-space trajectories). Second, it offers a library of composable operators, proximable functionals, and optimization algorithms, including a unified Fourier operator for all common trajectories and operators specifically developed for low-field applications, such as a B0-correction operator. These components are used to create ready-to-use implementations of key reconstruction algorithms. Third, for deep learning, MRpro includes essential building blocks such as data-consistency layers, differentiable optimization layers, state-of-the-art backbone networks, and access to public datasets to facilitate reproducibility. We demonstrate MRpro across automatic reconstruction, iterative SENSE, deep-learning-based reconstruction, and quantitative parameter estimation. Applications use public and simulated datasets and measured lowfield data acquired at 0.3 T, 0.6 T, and 47 mT.

eess.IV↗
arXiv2026

PyDoseRT Proton: A GPU Pencil-Beam Engine with a Convolutional Residual-Correction Network for Fast Proton Dose Calculation

Architecture category. Hybrid method: a physics-based analytical pencil-beam (PB) dose engine followed by a 3-D convolutional residual-correction network (RepVGG-U-Net). We addressed the DoseRAD2026 proton dose-prediction task with PyDoseRT Proton, a GPU-accelerated engine implemented in PyTorch and augmented by a learned residual toward Monte Carlo (MC) accuracy. A double-Gaussian PB kernel was calibrated to GATE/Geant4 integrated depth doses in water in two stages: a classical per-energy curve fit, then a gradient-based fit of the full 3-D dose through the PyTorch physics engine as it retains a differentiable execution path for gradient-based optimization of dose-dependent objectives. The engine computes each beamlet on a beam's-eye-view (BEV) lattice with variance-preserving Gaussian splitting, an analytic nuclear halo, and a Fermi-Eyges heterogeneity term, then rotates the result into the patient frame. Additionally, a compact residual U-Net predicts an additive correction in BEV space. It is conditioned on voxelwise material-label embeddings, a discrete energy embedding and spot size. The same model was used for all anatomical sites (thoracic and abdominal). It was trained with a patient-space L1 objective emphasizing the scored high-dose region and multi-scale BEV deep supervision. The submitted CT configuration obtained preliminary-test beamlet MAE 0.0066, image-z IDD distance 0.0025, plan MAE 0.0049, 98.30\% gamma pass rate (1\%/1 mm), and DVH error 0.460.

physics.med-ph↗
arXiv2026

Improving Clinical Target Volume Segmentation Accuracy using Anatomical Priors and Active Learning for the AGITG TOPGEAR Clinical Trial

Training deep learning-based medical image segmentation models is challenging with limited curated datasets. For AGITG TOPGEAR, a gastric cancer trial, the Clinical Target Volume (CTV) is complex and defined by multiple anatomical landmarks, making upfront training data preparation difficult for an automated contour QA segmentation model. We investigate anatomical priors, derived from surrounding organ segmentations, to provide spatial context and improve TOPGEAR CTV segmentation accuracy. We also evaluate active learning, iteratively expanding the training dataset by selecting cases expected to improve performance. One hundred TOPGEAR CT scans were retrospectively analyzed. An initial set of 10 expert-contoured cases was used to train an nnU-Net model. TotalSegmentator generated a voxel-wise anatomical prior map from surrounding structures as an additional input channel. Active learning was simulated over four iterations, selecting cases by model uncertainty and segmentation performance. All models used five-fold cross-validation for an ensemble uncertainty measure. Evaluation used a hold-out testing set of 50 cases. The anatomical prior improved CTV segmentation accuracy, increasing mean Dice Similarity Coefficient (DSC) from 0.84 to 0.86. Active learning similarly improved performance to 0.86, with greatest benefit in the final round. Combining the anatomical prior with active learning achieved the highest accuracy, with a DSC of 0.87. Model uncertainty correlated with DSC, supporting its use in identifying suboptimal predictions and guiding active learning. Anatomical priors and active learning each improved CTV segmentation accuracy and generalizability, with their combination achieving the best performance, supporting integration into segmentation model development for automated contour QA in radiotherapy clinical trials.

physics.med-ph↗