Related Experiment Video
Updated: Sep 3, 2025
Diffusion Imaging in the Rat Cervical Spinal Cord
Published on: April 7, 2015
Multiple B-Value Model-Based Residual Network (MORN) for Accelerated High-Resolution Diffusion-Weighted Imaging
Abstract:
Single-Shot Echo Planar Imaging (SSEPI) based Diffusion Weighted Imaging (DWI) has shortcomings such as low resolution and severe distortions. In contrast, Multi-Shot EPI (MSEPI) provides optimal spatial resolution but increases scan time. This study proposed a Multiple b-value mOdel-based Residual Network (MORN) model to reconstruct multiple b-value high-resolution DWI from undersampled k-space data simultaneously. We incorporated Parallel Imaging (PI) into a residual U-net to reconstruct multiple b-value multi-coil data with the supervision of MUltiplexed Sensitivity-Encoding (MUSE) reconstructed Multi-Shot DWI (MSDWI). Moreover, asymmetric concatenations among different b-values and the combined loss to back propagate helped the feature transfer. After training and validation of the MORN in a dataset of 32 healthy cases, additional assessments were performed on 6 patients with different tumor types. The experimental results demonstrated that the MORN model outperformed conventional PI reconstruction (i.e. SENSE) and two state-of-the-art deep learning methods (SENSE-GAN and VSNet) in terms of PSNR (Peak Signal-to-Noise Ratio), SSIM (Structual SIMilarity) and apparent diffusion coefficient maps. In addition, using the pre-trained model under DWI, the MORN achieved consistent fractional anisotrophy and mean diffusivity reconstructed from multiple diffusion directions. Hence, the proposed method shows potential in clinical application according to the observations on tumor patients as well as images of multiple diffusion directions.
Insights
A novel deep learning model, the Multiple b-value mOdel-based Residual Network (MORN), reconstructs high-resolution Diffusion Weighted Imaging (DWI) from undersampled data. This method improves image quality and shows promise for clinical applications in neuroimaging.
Area of Science:
- Magnetic Resonance Imaging (MRI)
- Medical Image Reconstruction
- Artificial Intelligence in Radiology
Background:
- Single-Shot Echo Planar Imaging (SSEPI) for Diffusion Weighted Imaging (DWI) suffers from low resolution and distortions.
- Multi-Shot EPI (MSEPI) offers better resolution but requires longer scan times.
- Efficient reconstruction of high-quality DWI is crucial for clinical diagnosis.
Purpose of the Study:
- To develop a deep learning model for simultaneous reconstruction of multiple b-value, high-resolution DWI from undersampled k-space data.
- To integrate Parallel Imaging (PI) with a residual U-net architecture for multi-coil data reconstruction.
- To validate the model's performance against existing reconstruction methods and in clinical cases.
Main Methods:
- Proposed a Multiple b-value mOdel-based Residual Network (MORN) incorporating PI and a residual U-net.
- Utilized MUltiplexed Sensitivity-Encoding (MUSE) reconstructed Multi-Shot DWI (MSDWI) for supervision.
- Employed asymmetric concatenations across b-values and a combined loss function for feature transfer.
- Trained and validated the MORN on 32 healthy cases and tested on 6 tumor patients.
Main Results:
- MORN significantly outperformed SENSE, SENSE-GAN, and VSNet in Peak Signal-to-Noise Ratio (PSNR) and Structural SIMilarity (SSIM).
- Reconstructed apparent diffusion coefficient (ADC) maps showed superior quality.
- Consistent fractional anisotropy (FA) and mean diffusivity (MD) were achieved using the pre-trained model for multiple diffusion directions.
Conclusions:
- The MORN model enables high-resolution, multiple b-value DWI reconstruction from undersampled data.
- The method demonstrates superior performance compared to conventional and deep learning-based techniques.
- MORN shows significant potential for clinical application in neuroimaging, particularly for tumor assessment.
