Variational Network with Wavelet-based UNET in Accelerated MRI Reconstruction from Under Sampled K-space Data
Researchers have developed a novel Variational Network incorporating a Wavelet-based U-Net (W-UNet) for accelerated MRI reconstruction. This method enhances the reconstruction of undersampled k-space data by replacing standard pooling operations with Discrete Wavelet Transform and Inverse Wavelet Transform modules. This approach preserves both low-frequency structures and high-frequency edge details, leading to improved artifact suppression and feature preservation. Experiments on the fastMRI knee and M4Raw brain datasets demonstrated state-of-the-art performance, validating the effectiveness of wavelet-based decomposition for accelerated MRI. AI