Abstract:Detail features of magnetic resonance images play a cru-cial role in accurate medical diagnosis and treatment, as they capture subtle changes that pose challenges for doc-tors when performing precise judgments. However, the widely utilized naive diffusion model has limitations, as it fails to accurately capture more intricate details. To en-hance the quality of MRI reconstruction, we propose a comprehensive detail-preserving reconstruction method using multiple diffusion models to extract structure and detail features in k-space domain instead of image do-main. Moreover, virtual binary modal masks are utilized to refine the range of values in k-space data through highly adaptive center windows, which allows the model to focus its attention more efficiently. Last but not least, an inverted pyramid structure is employed, where the top-down image information gradually decreases, ena-bling a cascade representation. The framework effective-ly represents multi-scale sampled data, taking into ac-count the sparsity of the inverted pyramid architecture, and utilizes cascade training data distribution to repre-sent multi-scale data. Through a step-by-step refinement approach, the method refines the approximation of de-tails. Finally, the proposed method was evaluated by con-ducting experiments on clinical and public datasets. The results demonstrate that the proposed method outper-forms other methods.

What problem does this paper attempt to address?

The problem this paper attempts to address is the accurate capture and preservation of detailed features in Magnetic Resonance Imaging (MRI) reconstruction. Specifically, existing simple diffusion models fail to accurately capture complex detailed information during the reconstruction process, leading to a decline in the quality of the reconstructed images and affecting doctors' ability to make precise diagnoses and treatments. To improve the quality of MRI reconstruction, the authors propose a detail-preserving reconstruction method based on multiple diffusion models (DP-MDM), aiming to enhance the preservation of image details by learning high-frequency information at multiple scales. ### Main Issues: 1. **Capture and Preservation of Detailed Features**: Existing diffusion models struggle to accurately capture and preserve complex detailed features, such as edges and textures, which are crucial for medical diagnosis and treatment. 2. **Multi-Scale Representation**: Current methods are insufficient in multi-scale representation, failing to effectively balance the capture of structural and detailed features. 3. **Handling Data Sparsity**: In the reconstruction of undersampled data, how to effectively handle data sparsity to improve reconstruction quality. ### Solutions: 1. **Multiple Diffusion Models**: Use multiple diffusion models to extract structural and detailed features in k-space, rather than extracting them in the image domain. 2. **Virtual Binary Modality Mask**: Utilize a virtual binary modality mask to refine the value range of k-space data, making the model more efficiently focus on detailed features through a highly adaptive central window. 3. **Inverted Pyramid Structure**: Construct an inverted pyramid structure to gradually reduce image information from top to bottom, achieving a multi-scale cascade representation of data. 4. **Cascaded Training Data Distribution**: Use cascaded training data distribution to represent multi-scale data, progressively refining the approximation of details. ### Experimental Validation: Experiments on clinical and public datasets validate the effectiveness of the proposed method. Results show that the DP-MDM method outperforms other methods in terms of detail preservation and image quality. ### Summary: This paper addresses the shortcomings of existing methods in capturing and preserving detailed features by proposing a detail-preserving reconstruction method based on multiple diffusion models, improving the quality of MRI reconstruction and aiding in the accuracy of medical diagnosis and treatment.

DP-MDM: Detail-Preserving MR Reconstruction via Multiple Diffusion Models

Rethinking Diffusion Model for Multi-Contrast MRI Super-Resolution

Discrete residual diffusion model for high-resolution prostate MRI synthesis

Correlated and Multi-frequency Diffusion Modeling for Highly Under-sampled MRI Reconstruction

Sub-DM:Subspace Diffusion Model with Orthogonal Decomposition for MRI Reconstruction

High-resolution MRI synthesis using a data-driven framework with denoising diffusion probabilistic modeling

MRI Super-Resolution with Partial Diffusion Models

MR Optimized Reconstruction of Simultaneous Multi-Slice Imaging Using Diffusion Model

Towards performant and reliable undersampled MR reconstruction via diffusion model sampling

Diffusion Modeling with Domain-conditioned Prior Guidance for Accelerated MRI and qMRI Reconstruction

High-Frequency Space Diffusion Models for Accelerated MRI

Score-based Diffusion Models With Self-supervised Learning For Accelerated 3D Multi-contrast Cardiac Magnetic Resonance Imaging

Arbitrary scale super-resolution diffusion model for brain MRI images

TC-DiffRecon: Texture coordination MRI reconstruction method based on diffusion model and modified MF-UNet method

Towards High-Quality MRI Reconstruction With Anisotropic Diffusion-Assisted Generative Adversarial Networks And Its Multi-Modal Images Extension

DisC-Diff: Disentangled Conditional Diffusion Model for Multi-Contrast MRI Super-Resolution

High-Frequency Space Diffusion Model for Accelerated MRI

Robust Physics-based Deep MRI Reconstruction Via Diffusion Purification

MRPD: Undersampled MRI reconstruction by prompting a large latent diffusion model

Zero-shot Dynamic MRI Reconstruction with Global-to-local Diffusion Model