Berkin Bilgic

Software:

Pulseq open-source pulse sequences:

Highly accelerated QSM acquisition using multi-echo GRE with B0 navigators: [GitHub]


(Siemens VE11) pulse sequences:

High resolution diffusion imaging with gSlider-SMS: [select "gSlider" from the drop-down list] [PDF]

Quantitative T1, T2 mapping at high resolution using 3D MR Fingerprinting: [select "3D MRF" from the drop-down list] [PDF]

High resolution, distortion-free diffusion imaging with BUDA-gSlider: [select "gSlider with BUDA" from the drop-down list] [PDF]


Machine learning:

Zero-shot Prior Learning of Spatio-temporal Multi-echo/contrast MRI Reconstruction with Iterative Refinement: [Jupyter Notebook] [PDF]

Quantifying the uncertainty of neural networks using Monte Carlo dropout for safer and more accurate deep learning based quantitative MRI: [Python code] [ISMRM'22 abstract]

MAPLE: Accelerated MR Parameter Mapping with Scan-specific Unsupervised Networks: [Python code] [ISMRM'22 abstract]

Latent Signal Modeling (LSM): Learning compact latent representations of signal evolution for improved shuffling reconstruction: [Python code] [ISMRM'22 abstract]

Rapid quantitative imaging with Wave-MODL and joint reconstruction: [Python code] [ISMRM'22 abstract]

Wave-MODL: Wave-Encoded Model-Based Deep Learning: [Python code] [ISMRM abstract]

eRAKI: Fast Robust Artificial neural networks for Kā€space Interpolation: [Python code] [ISMRM abstract]

Scan-specific, Parameter-free Artifact Reduction in K-space (SPARK): [Python code for the ISMRM abstract] [ISMRM abstract]
[Python code for the MRM paper] [MRM preprint]

Accelerated Multi-shot EPI through Machine Learning and Joint Reconstruction: [Data & Python code for Keras] [PDF]


Distortion-free Echo Planar Imaging:

VUDU-SAGE: Efficient T2 and T2* Mapping using Joint Reconstruction for Motion-Robust, Distortion-Free, Multi-Shot, Multi-Echo EPI: [Matlab code] [PDF]

VUDU: motion-robust, distortion-free multi-shot EPI: [ISMRM'22 Github code] [ISMRM'22 abstract]
[ISMRM'21 Matlab code] [ISMRM'21 abstract]

T2-BUDA-gSlider: rapid high resolution T2 mapping with blip-up/down acquisition, gSlider and subspace reconstruction: [Matlab code] [preprint]


Parallel imaging and compressed sensing:

VC-MUSSELS: Hankel low-rank regularization for multi-shot EPI with virtual coils for high-fidelity partial Fourier reconstruction: [HTML] [Matlab code]

Calibrationless parallel imaging for multi echo/contrast data: [Matlab code] [HTML]

Joint Reconstruction:

Wave-CAIPI:


Receiver coil combination without reference data:

Block Coil Compression (BCC) for reference-free coil combination at ultra high field: [Matlab code] [PDF]

SVD for reference-free coil combination at high field: [Matlab code] [HTML]


Quantitative Susceptibility Mapping (QSM):

Dipole Inversion Algorithms:

Single-Step QSM:

Demos and Recon Challenges:


MR Spectroscopic Imaging:

Fast lipid suppression with l2-regularization: [Matlab code] [PDF]

Lipid suppression with spatial priors and l1-regularization: [Matlab code] [PDF]


Accelerated Diffusion Spectrum Imaging:

Fast Diffusion Spectrum Imaging reconstruction with trained dictionaries: [Matlab code] [PDF]

Diffusion Spectrum Imaging with dictionary reconstruction: [Matlab code] [PDF]

Contact Information:

Berkin

Room 2.102, Building 75
13th Street
Charlestown, MA 02129

E-mail: berkin AT nmr.mgh.harvard.edu

Tel: 617-866-8740