mlsanityData to 7 Oct 2026

Tracing the Oracle: Improving Diffusion Timestep Scheduling for 3D CT Reconstruction

arXiv 2606.06236 ↗First posted 4 Jun 2026cs.LGcited by 0

On the map: Diffusion models → Medical imaging and segmentation

Abstract (arXiv)

Pretrained diffusion models demonstrate impressive potential in solving highly ill-posed 3D computed tomography (CT) inverse problems, while the inference process suffers from significant computational overhead. Furthermore, existing uniform timestep schedules fail to capture the non-uniform evolution of the reverse conditional diffusion stochastic differential equation, thereby introducing substantial truncation errors. To overcome this limitation, we propose Tracing the Oracle (TrO), a plug-and-play framework for improved timestep scheduling. Specifically, we treat densely sampled numerical integration trajectories on a few samples as the reference oracle. The optimized schedule is extracted by leveraging dynamic programming to globally minimize the cumulative error between the few-step approximation and the oracle. This mechanism precisely allocates the limited sampling steps to critical evolution stages that are highly susceptible to truncation errors. Our extensive experiments on the AAPM dataset across multiple 3D CT reconstruction tasks demonstrate that, when combined with the state-of-the-art 3D CT reconstruction method DDS, our optimized timesteps significantly improve reconstruction fidelity and computational efficiency compared to existing heuristic schedules, especially under a strict budget of no more than 10 sampling steps.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models. Builds on 35 of its 45 references that are in the map. First posted on arXiv 4 Jun 2026.

Builds on

35 of 45 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024DiffusionBlend: Learning 3D Image Prior through Position-aware Diffusion Score Blending for 3D Computed Tomography Reconstruction14 Jun
  2. 2022Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models19 Nov
  3. 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Models15 Nov

Other references · 32

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Denoising Diffusion Implicit Modelscited by 9,655
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  4. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 2,994
  5. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Stepscited by 1,961
  6. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,375
  7. 2014Image Super-Resolution Using Deep Convolutional Networkscited by 1,218
  8. 2021Multi-Stage Progressive Image Restorationcited by 813
  9. 2022Improving Diffusion Models for Inverse Problems using Manifold Constraintscited by 501
  10. 2023UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Modelscited by 441
  11. 2017Compressed Sensing using Generative Modelscited by 337
  12. 2023Denoising Diffusion Models for Plug-and-Play Image Restorationcited by 335
20 more
  1. 2023Inversion by Direct Iteration: An Alternative to Denoising Diffusion for Image Restorationcited by 164
  2. 2024Improving Diffusion Inverse Problem Solving with Decoupled Noise Annealingcited by 157
  3. 2023Decomposed Diffusion Sampler for Accelerating Large-Scale Inverse Problemscited by 129
  4. 2024Align Your Steps: Optimizing Sampling Schedules in Diffusion Modelscited by 91
  5. 2023Fast ODE-based Sampling for Diffusion Models in Around 5 Stepscited by 83
  6. 2021Intermediate Layer Optimization for Inverse Problems using Deep Generative Modelscited by 70
  7. 2024Accelerating Diffusion Sampling with Optimized Time Stepscited by 67
  8. 2024On the Trajectory Regularity of ODE-based Diffusion Samplingcited by 51
  9. 2024Improving Diffusion Models for Inverse Problems Using Optimal Posterior Covariancecited by 42
  10. 2025S4S: Solving for a Diffusion Model Solvercited by 10
  11. 2025Learnable Sampler Distillation for Discrete Diffusion Modelscited by 8
  12. 2025Provable Diffusion Posterior Sampling for Bayesian Inversioncited by 8
  13. 2022Misspecified Phase Retrieval with Generative Priorscited by 7
  14. 2023A Unified Framework for Uniform Signal Recovery in Nonlinear Generative Compressed Sensingcited by 7
  15. 2024Unleashing the Denoising Capability of Diffusion Prior for Solving Inverse Problemscited by 7
  16. 2025Integrating Intermediate Layer Optimization and Projected Gradient Descent for Solving Inverse Problems with Diffusion Modelscited by 7
  17. 2023Solving Quadratic Systems with Full-Rank Matrices Using Sparse or Generative Priorscited by 4
  18. 2026Just-in-Time: Training-Free Spatial Acceleration for Diffusion Transformerscited by 4
  19. 2025Image Restoration via Primal Dual Hybrid Gradient and Flow Generative Modelcited by 2
  20. 2026Image Restoration via Diffusion Models with Dynamic Resolutioncited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026DM4CT: Benchmarking Diffusion Models for Computed Tomography Reconstruction
  2. 2025Local Patches Meet Global Context: Scalable 3D Diffusion Priors for Computed Tomography Reconstruction
  3. 2026Timestep Rescheduling in Diffusion Inversion
  4. 2026Efficient Flow Matching for Sparse-View CT Reconstruction
  5. 2025NERD: Network-Regularized Diffusion Sampling For 3D Computed Tomography
  6. 2025When are Diffusion Priors Helpful in Sparse Reconstruction? A Study with Sparse-view CT
  7. 2025Diffusion Active Learning: Towards Data-Driven Experimental Design in Computed Tomography
  8. 2026From Sparse X-rays to 3D CT: Training-Free Reconstruction with Diffusion Priors