AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Masked-Diffusion Autoencoders for 3D Medical Vision Representation Learning

Jiachen Tu, Guanghui Qin, Theodore Zhengde Zhao, Jeya Maria Jose Valanarasu, Sheng Zhang, Tristan Naumann, Fan Lam, Sheng Wang, Hoifung Poon

CVPR 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training

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
combine spatial masking with diffusion noise (dual corruption) (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: 0.598 -> 0.658Improved
sample masking ratio from U(0.01, 0.99) instead of a fixed ratio (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: 0.625 -> 0.662Improved
raise fixed masking ratio from 75% to 95% (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: 0.625 -> 0.568Hurt
use block-wise 16^3 masking instead of random voxel masking (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: 0.600 -> 0.623Improved
corrupt raw pixel space instead of embedding space (EMDAE) (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: +4.2-10.2% AUROC over EMDAEImproved
switch noise schedule among VP, VE and flow matching (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: VE 0.615, Flow Matching 0.614, VP 0.613No clear effect
use ResNet encoder instead of Vision Transformer (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROC: +12.1-13.8% AUROC over ViTImproved
add FiLM time conditioning to ResNet (ablation)MDAE · BraTS18 LGG vs HGG ablationAUROCNo clear effect
add masked reconstruction loss alongside visible denoising loss (ablation)MDAE · UCSF-PDGM segmentation (mean Dice)Dice (%): 91.8 -> 92.6Improved
add visible-region denoising loss alongside masked reconstruction loss (ablation)MDAE · UCSF-PDGM classification (mean)classification mean score (%): 86.4 -> 90.7Improved
pretrain with MDAE instead of MAEResNet U-Net encoder · Cross-modal OOD brain MRI (6 tasks)mean AUROC (%): 70.0 -> 78.6Improved
pretrain with MDAE instead of MAEResNet U-Net encoder · In-distribution brain MRI (T1/T2, 6 tasks)mean AUROC (%): 69.5 -> 73.6Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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