Deep MRI Synthesis Collection

Overview

This is the official collection of deep learnning based MRI synthesis methods. Included a various range of MRI modalities including T1 weighted MRI, diffusion weighted MRI (dMRI), Diffusion Tensor Image (DTI), and Diffusion Fiber Orientation Distribution (FOD).

overview

Our Works

  • Enhancing Angular Resolution via Directionality Encoding and Geometric Constraints in Brain Diffusion Tensor Imaging

    Sheng Chen, Zihao Tang, Mariano Cabezas, Xinyi Wang*, Chenyu Wang, Weidong Cai.

       [Publication]                    [Code]

  • Enhancing Angular Resolution via Directionality Encoding and Geometric Constraints in Brain Diffusion Tensor Imaging

    Sheng Chen, Zihao Tang, Mariano Cabezas, Xinyi Wang*, Arkiev D’Souza, Michael Barnett, Fernando Calamante, Weidong Cai, Chenyu Wang.

       [Publication]                    [Code]

  • High Angular Diffusion Tensor Imaging Estimation from Minimal Evenly Distributed Diffusion Gradient Directions

    Zihao Tang, Sheng Chen, Arkiev D’Souza, Dongnan Liu, Fernando Calamante, Michael Barnett, Weidong Cai, Chenyu Wang, and Mariano Cabezas.

       [Publication]                    [Code]

  • TW-BAG: Tensor-wise Brain-aware Gate Network for Inpainting Disrupted Diffusion Tensor Imaging (DICTA2022)

    Zihao Tang*, Xinyi Wang*, Lihaowen Zhu, Mariano Cabezas, Dongnan Liu, Michael Barnett, Weidong Cai, and Chenyu Wang.

       [Publication]                    [Code]

  • Diffusion MRI Fibre Orientation Distribution Inpainting (MICCAI2022 CDMRI Workshop)

    Zihao Tang*, Xinyi Wang*, Mariano Cabezas, Arkiev D’Souza, Fernando Calamante, Dongnan Liu, Michael Barnett, Chenyu Wang, and Weidong Cai.

       [Publication]                    [Code]

  • LG-Net: Lesion Gate Network for Multiple Sclerosis Lesion Inpainting (MICCAI2021)

    Zihao Tang, Mariano Cabezas, Dongnan Liu, Michael Barnett, Weidong Cai, and Chenyu Wang.

       [Publication]                    [Code]

  • Highlights

    OCE-Net (Diffusion MRI Fibre Orientation Distribution Inpainting)

    • Present a Order-wise Coefficient Estimation Network to inpaint the disrupted Fibre Orientation Distribution (FOD) coefficients.
    • Order-wise Coefficient Decoders to decode the coefficient independently.
    • Explore the analysis on connectome.

    LG-Net

    • Present an end-to-end solution for Multiple Sclerosis (MS) lesion inpainting on T1 MRI images.
    • A specifically designed lesion gate consistency.
    • State-of-the-art on computer vision similaritiy metrics and brain tissue volumetric metrics.

    BibTeX

    If you find our data or project useful in your research, please cite:
    @inproceedings{tang2021lg,
        title="LG-Net: Lesion Gate Network for Multiple Sclerosis Lesion Inpainting",
        author="Tang, Zihao and Cabezas, Mariano and Liu, Dongnan and Barnett, Michael and Cai, Weidong and Wang, Chenyu",
        booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)},
        pages="660--669",
        year={2021},
        organization={Springer}
    }
                    

    Acknowledgments

    The website template was borrowed from Chaoyi Zhang.