Single Image Compressed Sensing MRI via a Self-Supervised Deep Denoising Approach
This work addresses the challenge of generalizability and data access in compressed sensing MRI for real-world applications, offering an incremental improvement over existing methods.
The paper tackles the problem of compressed sensing MRI reconstruction by proposing a single image, self-supervised framework that combines deep and sparse regularization to dampen structured CS artefacts, resulting in an average PSNR improvement of 2-4dB on brain and knee datasets.
Popular methods in compressed sensing (CS) are dependent on deep learning (DL), where large amounts of data are used to train non-linear reconstruction models. However, ensuring generalisability over and access to multiple datasets is challenging to realise for real-world applications. To address these concerns, this paper proposes a single image, self-supervised (SS) CS-MRI framework that enables a joint deep and sparse regularisation of CS artefacts. The approach effectively dampens structured CS artefacts, which can be difficult to remove assuming sparse reconstruction, or relying solely on the inductive biases of CNN to produce noise-free images. Image quality is thereby improved compared to either approach alone. Metrics are evaluated using Cartesian 1D masks on a brain and knee dataset, with PSNR improving by 2-4dB on average.