AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DM4CT: Benchmarking Diffusion Models for Computed Tomography Reconstruction

Jiayang Shi, Daniel Pelt, Joost Batenburg

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Diffusion posterior sampling · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use Poisson-noise measurements instead of noise-free onesDDS · Medical CT (config i vs ii)PSNR (dB): 31.43 -> 20.12Hurt
use Poisson-noise measurements instead of noise-free onesDDS · Industrial CT (config i vs ii)PSNR (dB): 22.87 -> 18.23Hurt
replace Poisson noise with matched-level Gaussian noise (ablation)DDS · Noise-model comparisonreconstruction qualityImproved
increase DC step size eta moderately (ablation)DPS · CT reconstructionPSNR and data fitImproved
increase DC step size eta too far (ablation)DPS · CT reconstructionPSNRHurt
add explicit data-consistency optimization steps to latent diffusion (ReSample vs gradient-only PSLD) (ablation)Latent diffusion · Medical CT noise-free (40 angles)reconstruction coherenceImproved
add explicit data-consistency optimization steps to latent diffusion under noise (ablation)ReSample · Medical CT noisy (80 angles)image qualityHurt
use supervised training (SwinIR) instead of diffusion priorsDiffusion methods · Medical CT config ivPSNR (dB): 29.70 -> 30.79Improved
use diffusion priors instead of classical FBP/SIRTClassical reconstruction · Medical CT config ivPSNR (dB): 25.86 -> 29.70Improved
use pixel-space instead of latent-space diffusionDiffusion methods · CT reconstruction efficiencytime and memoryImproved
use diffusion priors under sparse/high-noise measurementsClassical and MBIR methods · Medical and industrial CTPSNRImproved
apply empirical linear value-range mapping to raw projectionsMCG · Medical CT raw projectionsPSNR (dB): 19.48 -> 29.12Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy