TY - GEN
T1 - Learned Primal Dual Splitting for Self-Supervised Noise-Adaptive MRI Reconstruction
AU - Janjušević, Nikola
AU - Khalilian-Gourtani, Amirhossein
AU - Wang, Yao
AU - Feng, Li
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Magnetic resonance imaging (MRI) reconstruction has largely been dominated by deep neural networks (DNN); however, many state-of-the-art architectures use black-box structures, which hinder interpretability and improvement. Here, we propose an interpretable DNN architecture for self-supervised MRI reconstruction and denoising by directly parameterizing and learning the classical primal-dual splitting, dubbed LPDSNet. This splitting algorithm allows us to decouple the observation model from the signal prior. Experimentally, we show other interpretable architectures without this decoupling property exhibit failure in the self-supervised learning regime. We report state-of-the-art self-supervised joint MRI reconstruction and denoising performance and novel noise-level generalization capabilities, where in contrast black-box networks fail to generalize.
AB - Magnetic resonance imaging (MRI) reconstruction has largely been dominated by deep neural networks (DNN); however, many state-of-the-art architectures use black-box structures, which hinder interpretability and improvement. Here, we propose an interpretable DNN architecture for self-supervised MRI reconstruction and denoising by directly parameterizing and learning the classical primal-dual splitting, dubbed LPDSNet. This splitting algorithm allows us to decouple the observation model from the signal prior. Experimentally, we show other interpretable architectures without this decoupling property exhibit failure in the self-supervised learning regime. We report state-of-the-art self-supervised joint MRI reconstruction and denoising performance and novel noise-level generalization capabilities, where in contrast black-box networks fail to generalize.
KW - Deep-learning
KW - interpretable neural network
KW - joint MRI reconstruction and denoising
KW - unrolled network
UR - https://www.scopus.com/pages/publications/105005826809
U2 - 10.1109/ISBI60581.2025.10980799
DO - 10.1109/ISBI60581.2025.10980799
M3 - Conference contribution
AN - SCOPUS:105005826809
T3 - Proceedings - International Symposium on Biomedical Imaging
BT - ISBI 2025 - 2025 IEEE 22nd International Symposium on Biomedical Imaging, Proceedings
PB - IEEE Computer Society
T2 - 22nd IEEE International Symposium on Biomedical Imaging, ISBI 2025
Y2 - 14 April 2025 through 17 April 2025
ER -