Uncertainty-Propelled Physics-MAE Fusion for Self-Supervised Diffusion-Weighted Image Denoising
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (5) · 3578-3586
DOI: 10.1609/aaai.v40i5.37356 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Zeyu Deng und 6 weiteren Autor:innen.
Abstract (Originalsprache)
The inherently low signal-to-noise ratio (SNR) in diffusion-weighted (DW) imaging fundamentally impedes precise tissue microstructure characterization, rendering effective noise suppression a persistent challenge. Existing denoising methods frequently suffer from over-smoothing or distortion of microstructure information when handling spatially correlated or severe noise. To address these limitations, we propose UP2-MAE fusion model, a self-supervised DWI denoising method based on Uncertainty-Propelled Physics and Masked Auto-Encoder (MAE) fusion. This framework integrates two complementary branches: one leverages MAE to suppress noise through local context modeling, while the other constructs uncorrelated noisy pairs using diffusion tensor imaging (DTI) physics and denoises them via a Noise2Noise approach, which can preserve texture details by exploiting directional relationships across diffusion encoding directions. To fully integrate the strengths of both branches, an uncertainty-propelled fusion strategy based on maximum likelihood estimation is proposed to derive the final denoised output. In addition, to further promote the performance, uncertainty-guided reconstruction and consistency loss are presented. Evaluations against state-of-the-art denoising methods on both simulated and acquired DW datasets confirm the efficacy of our approach.
Zitieren
APA
Deng, Z., Wang, L., Tao, X., Chen, Q., Cao, Y., XuLinHu, & Ou, Y. (2026). Uncertainty-Propelled Physics-MAE Fusion for Self-Supervised Diffusion-Weighted Image Denoising. Proceedings of the AAAI Conference on Artificial Intelligence, 40(5), 3578-3586. https://doi.org/10.1609/aaai.v40i5.37356
Vancouver
Deng Z, Wang L, Tao X, Chen Q, Cao Y, XuLinHu, et al. Uncertainty-Propelled Physics-MAE Fusion for Self-Supervised Diffusion-Weighted Image Denoising. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(5):3578-3586. doi:10.1609/aaai.v40i5.37356
BibTeX
@article{Deng2026,
author = {Deng, Zeyu and Wang, Lihui and Tao, Xi and Chen, Qijian and Cao, Ying and XuLinHu and Ou, Yingfeng},
title = {Uncertainty-Propelled Physics-MAE Fusion for Self-Supervised Diffusion-Weighted Image Denoising},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {5},
pages = {3578--3586},
doi = {10.1609/aaai.v40i5.37356}
}
EUVIMED