Software¶

MATLAB QSM reconstruction pipelines for different MRI scanners and sequences, including single-echo and multi-echo GRE and EPI and ME-MP2RAGE.

RESHARP (Regularization Enabled SHARP) is a background field removal method widely adopted in Quantitative Susceptibility Mapping (QSM).

POEM (Phase-Offsets Estimation Method) is an Multi-echo arrayed coil combination for phase method, enables accurate coil combination for multi-echo acquisitions, improving SNR and phase consistency.

A collection of deep learning-based MRI methods, such as QSM and undersampling reconstruction and MRF recon.

Octave-convolutional neural network for dipole inversion. xQSM uses octave convolutions and noise-regularization to robustly solve the dipole inversion problem in QSM.

Single-step deep learning QSM from raw phase. iQSM is a deep learning method that directly reconstructs susceptibility maps from raw phase, bypassing multi-step processing.

Extended iQSM method for single-step QSM. iQSM+ builds on iQSM with improved network design and robustness, enabling accurate susceptibility maps even with challenging acquisitions.

ODCRNet: Accelerating Quantitative Susceptibility Mapping using Compressed Sensing and Deep Neural Network. A deep complex residual network (DCRNet) to recover both MR magnitude and quantitative phase images from the CS undersample k-space data, enabling the acceleration of QSM acquisitions.

AFTER-QSM for acquisition of any fov orientations (beyond pure axial). AFTER-QSM uses affine transformation to handle scans of angled FOV, improving robustness and generalizability.

Model-based Deep Image Prior for QSM. MoDIP is an unsupervised dipole inversion method leveraging deep image prior principles for quantitative susceptibility mapping without training data.

Fast Controllable Diffusion Models for Undersampled MRI Reconstruction. This study introduces a new algorithm called Predictor-Projector-Noisor (PPN), which enhances and accelerates controllable generation of diffusion models for undersampled MRI reconstruction. .

Diffusion-model-based QSM for EPI scans. QSMDiff employs diffusion probabilistic models to reconstruct QSM from undersampled or distorted EPI scans, improving robustness and generalizability.