mlsanityData to 14 Oct 2026

Unsupervised Anatomical Feature Learning via Diffusion Models: Enhanced Medical Image Segmentation with Denoising Diffusion Probabilistic Models

arXiv 2608.25693 ↗First posted 26 Aug 2026cs.CV cs.AI cs.LGcited by 0

On the map: Diffusion models → Medical imaging and segmentation

Abstract (arXiv)

Acquiring pixel-level annotations for medical image segmentation is a severe bottleneck. Traditional U-Net architectures, while effective, learn local texture patterns and lack awareness of global anatomical structures, leading to boundary delineation failures in low-data regimes. This research paper proposes utilizing unsupervised Denoising Diffusion Probabilistic Models (DDPMs) to extract anatomical features. We train a DDPM on 21 unlabeled abdominal CT scans to learn structural representations, transferring the encoder weights to a downstream segmentation task evaluated on the BTCV multi-organ dataset. Diffusion pretraining significantly improved liver segmentation: Dice increased from 0.75±0.36 to 0.93±0.16 (p < 5.33×10-26, 0.529 Cohen's d), Average Surface Distance (ASD) decreased by 66%, and 95th-percentile Hausdorff Distance (HD95) reduced by 45%. For kidney segmentation, Dice improved from 0.90±0.19 to 0.95±0.10 (p < 4.01×10-11). Multi-organ pooled performance showed a 68% variance reduction and a 74% improvement in boundary precision (Dice 0.95±0.07). Crucially, frozen encoder models retained > 80% of fine-tuned performance without exposure to segmentation labels, proving the existence of learned anatomical priors. In low-data scenarios, diffusion-pretrained models maintained robust performance with only 50% (Dice: 0.92 liver, 0.94 kidney), 25%, and even 10% (Dice: 0.89 liver, 0.71 kidney) of labeled data. Using unlabeled images for diffusion-based pretraining successfully embeds robust anatomical features prior to human supervision, transforming U-Nets into anatomy-aware systems.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models. Builds on 35 of its 62 references that are in the map. First posted on arXiv 26 Aug 2026.

Builds on

35 of 62 references.

Closest ancestors · 8

Same direction, newest first

  1. 2023GenSelfDiff-HIS: Generative Self-Supervision Using Diffusion for Histopathological Image Segmentation4 Sep
  2. 2022Diffusion Probabilistic Models beat GANs on Medical Images14 Dec
  3. 2022Diffusion Models for Medical Image Analysis: A Comprehensive Survey14 Nov
  4. 2022MedSegDiff: Medical Image Segmentation with Diffusion Probabilistic Model1 Nov
  5. 2022PET image denoising based on denoising diffusion probabilistic models13 Sep
  6. 2021Label-Efficient Semantic Segmentation with Diffusion Models6 Dec
  7. 2021SegDiff: Image Segmentation with Diffusion Probabilistic Models1 Dec
  8. 2021Score-based diffusion models for accelerated MRI8 Oct

Other references · 27

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  5. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  6. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  7. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  8. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,611
  9. 2016V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentationcited by 2,292
  10. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
  11. 2017A Survey on Deep Learning in Medical Image Analysiscited by 1,186
  12. 2022Diffusion Models in Vision: A Surveycited by 1,015
15 more
  1. 2019Self-supervised Visual Feature Learning with Deep Neural Networks: A Surveycited by 657
  2. 2019UNet++: Redesigning Skip Connections to Exploit Multiscale Features in Image Segmentationcited by 586
  3. 2024Diffusion Model-Based Image Editing: A Surveycited by 212
  4. 2023A Survey on Self-supervised Learning: Algorithms, Applications, and Future Trendscited by 207
  5. 2019MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentationcited by 195
  6. 2022Medical Image Segmentation Review: The success of U-Netcited by 191
  7. 2020U-Net and its variants for medical image segmentation: theory and applicationscited by 169
  8. 2021DS-TransUNet:Dual Swin Transformer U-Net for Medical Image Segmentationcited by 126
  9. 2020CA-Net: Comprehensive Attention Convolutional Neural Networks for Explainable Medical Image Segmentationcited by 79
  10. 2023AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weekscited by 52
  11. 2020Residual Attention U-Net for Automated Multi-Class Segmentation of COVID-19 Chest CT Imagescited by 24
  12. 2023Probabilistic Modeling of Inter- and Intra-observer Variability in Medical Image Segmentationcited by 17
  13. 2025Calibration and Uncertainty for multiRater Volume Assessment in multiorgan Segmentation (CURVAS) challenge resultscited by 7
  14. 2023Diffusion Models for Generative Artificial Intelligence: An Introduction for Applied Mathematicianscited by 5
  15. 2023How inter-rater variability relates to aleatoric and epistemic uncertainty: a case study with deep learning-based paraspinal muscle segmentationcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Label-efficient multi-organ segmentation with a diffusion model
  2. 2025Medical Semantic Segmentation with Diffusion Pretrain
  3. 2024Latent Diffusion for Medical Image Segmentation: End to end learning for fast sampling and accuracy
  4. 2024Enhancing Medical Image Segmentation with Deep Learning and Diffusion Models
  5. 2023A Recycling Training Strategy for Medical Image Segmentation with Diffusion Denoising Models
  6. 2024Analysing Diffusion Segmentation for Medical Images
  7. 2023Diff-UNet: A Diffusion Embedded Network for Volumetric Segmentation
  8. 2023Importance of Aligning Training Strategy with Evaluation for Diffusion Models in 3D Multiclass Segmentation