Advanced MRI Techniques and Applications

This paper improves the robustness of 3D multi-shot EPI for fMRI at 7T by combining a new sampling trajectory with a structured low-rank reconstruction method, achieving improved tSNR and acceleration.

This paper develops a two-stage deep learning model to reduce motion artifacts in liver DCE-MRI, achieving better results than existing methods.

The paper develops a mechanistic model integrating neurovascular and neurometabolic couplings to describe and predict metabolic and BOLD responses to visual stimuli.

This paper optimizes a non-contrast MRI technique to generate 4D angiographic and perfusion images of the whole head from a single scan using radial imaging and arterial spin labeling.

The paper introduces DANSE, a deep learning method for fast and noise-suppressed estimation of qGRE MRI metrics, outperforming traditional methods in speed and accuracy.

HALO is a software tool that provides real-time head alignment feedback to improve consistency across MRI sessions, reducing susceptibility noise in longitudinal studies.

The paper provides a high-resolution T2* MRI dataset of the human brain at 7 Tesla, revealing cortical substructures and consistent depth-dependent T2* values across subjects.

This paper develops unsupervised deep learning methods for frequency and phase correction of MRS signals, showing comparable performance to supervised approaches and improved efficiency.

The paper investigates neurometabolic changes with age using MR spectroscopy in a cross-sectional cohort, finding age correlations for several metabolites but not others.

This paper introduces TQ-SILiCON, a method that fuses quantitative susceptibility maps with T1-weighted images to improve brain tissue contrast in primates, enhancing segmentation of subcortical and white matter structures.

This paper conducts a systematic review and meta-analysis of fMRS studies on GABA, Glu, and Glx, finding small to moderate effect sizes for Glu and Glx but no significant effects for GABA.

This study evaluates ME-ICA denoising on task-based fMRI, finding it improves model fitting performance compared to single echo and optimal combination without ICA.

The paper measures cardiac-aligned fast fMRI signals to assess blood and CSF flow dynamics in the brain using SMS acquisition methods.

The paper links quantitative MRI parameters to cell type-specific gene expression in human neocortex, showing associations with GABA- and glutamatergic neurons and other cell types.

The paper evaluates and corrects slice-direction geometric distortion in high-resolution 7T EPI using reversed slice-select gradient acquisitions and gradient reversal-based correction.

GLMsingle improves single-trial fMRI response estimates using custom HRFs, noise regressors, and ridge regression, enhancing reliability and decoding performance across datasets.

The paper develops a model to predict depth-dependent VASO and BOLD responses in human primary visual cortex and finds that CBV changes mainly occur in small arterioles and capillaries.

The paper investigates the relationship between cerebral blood flow (CBF) and negative BOLD responses (NBR) in the human visual cortex, comparing them to positive BOLD responses (PBR) using multi-echo imaging techniques.

The paper introduces NIfTI-MRS, a standardized format for magnetic resonance spectroscopy data to improve sharing and integration across the MRS community.

This paper presents quantitative metabolite maps of 12 metabolites in human brains at 9.4 T, using voxel-specific T1-relaxation corrections for water and metabolites.

This paper introduces a goal-specific MRI harmonization framework that improves downstream prediction performance by integrating application-specific constraints during harmonization.

The paper investigates non-neural factors affecting BOLD response magnitudes in human participants and proposes a correction method to improve fMRI data interpretation.

The paper introduces a method to estimate local errors in multi-parameter mapping (MPM) neuroimaging data, improving group-level sensitivity by accounting for noise and artefacts.

This paper quantifies cerebral blood arrival times using hypoxia-mediated arterial BOLD contrast and applies it to assess collateral flow in patients.

This paper demonstrates that 2x zero-filling improves modeling accuracy in in vivo MRS data compared to no zero-filling, with no additional benefit from higher zero-filling.

This study compares EPImix and conventional T1-weighted MRI scans for tissue volume estimation and brain-age prediction, finding similar results but a systematic age offset that can be corrected.

The paper shows visual stimulation induces fast, functionally specific cortical thickness changes in V1, comparable to BOLD signals but weaker in V3, suggesting non-hemodynamic mechanisms.

This study compares multi-echo combination methods for MRI quantitative susceptibility mapping (QSM) and finds that nonlinear phase fitting before Laplacian-based methods provides the highest accuracy and lowest noise.

This study evaluates frame censoring for motion correction in task-based fMRI across multiple datasets, finding modest improvements but no single method consistently outperforms others.