Advanced MRI Techniques and Applications

The paper tests lactate shuttling in the human brain using hyperpolarized 13C-MRI and finds that reducing lactate saturation increases bicarbonate signal, indicating lactate conversion to pyruvate and oxidation.

This paper assesses how scanner motion affects brain age gap estimates using MRI data from two datasets, finding significant differences in predicted age with varying motion levels.

This study evaluates high-resolution arterial spin labelling MRI at 3 Tesla, showing improved SNR and reduced partial voluming compared to standard protocols.

This paper investigates how anatomical and vascular features of the human brain affect the temporal specificity of fMRI signals in the primary visual cortex.

This study develops a ME-EPI protocol for olfactory task-based fMRI, showing improved BOLD sensitivity over 1E-EPI through signal combination and artifact reduction.

This study evaluates the reliability of task-based fMRI in the human spinal cord, finding poor reliability of BOLD responses across days but good reliability of autonomic responses.

This study shows that cerebrovascular reactivity increases with age across multiple brain networks using longitudinal fMRI data from children.

This study demonstrates the feasibility of BOLD fMRI in detecting neuronal activation in the lumbosacral cord during lower extremity motor tasks and shows TE-dependent effects on BOLD signal.

This study evaluates two versions of NORDIC denoising on submillimeter auditory fMRI data, finding improved detection sensitivity and reliability but reduced average response amplitude.

The paper proposes methods to approximate bone mineral density and subcutaneous adiposity from T1-weighted head MRI images and validates their reliability and correlation with DXA and MRI measurements.

The paper introduces BrainQCNet, a deep learning model for automated artifact detection in brain structural MRI scans, achieving high accuracy and fast processing times.

This paper investigates noise in line-scan MRI and shows that temporal fluctuations can mimic neural responses, highlighting limitations in detecting the DIANA response.

The paper introduces FMRI quality control tools in AFNI, providing systematic, in-depth, and interactive QC features for preprocessing pipelines.

This paper characterizes cortical depth-dependent CBV and BOLD hemodynamic responses across stimulus durations using VASO and ME-GRE at 7T in humans, finding that longer durations improve CBV specificity but may reduce efficiency.

This study evaluates how acquisition and modeling parameters affect test-retest reproducibility of GABA+ measurements using edited MRS, finding that model selection significantly impacts reproducibility.

This study evaluates ME-ICA's effectiveness in reducing task-correlated head motion artifacts in motor-task fMRI data, showing improved signal clarity and stability in both subject and group analyses.

This study assesses the intra-individual reliability of 1H-MRS measurements in the anterior cingulate cortex over one year, finding that metabolite levels are reliably quantified with small variability.

The paper develops a resource to evaluate MRI harmonisation approaches by comparing variability across scanners, within-scanners, and biological differences using multimodal data from multiple sites.

The paper examines age-related changes in basal ganglia iron levels in neonates using rsfMRI data, finding postnatal age associations in specific subregions and preterm birth effects.

This study uses interleaved fMRS and fMRI to examine how visual contrast affects neurometabolite levels and BOLD responses in the human brain, finding non-linear relationships and consistent glutamate increases.

This paper demonstrates a method to concurrently measure BOLD and metabolite changes during a cognitive task using fMRS, showing significant Glx increases and moderate BOLD correlation.

This paper introduces AGNOSTIC, a synthetic 1H MRS dataset for training deep learning models to detect and predict out-of-voxel echoes in brain MRS data.

The paper develops acquisition protocols and processing pipelines for whole-brain layer-fMRI and introduces the Kenshu dataset for benchmarking and analysis.

This paper characterizes cardiac-induced noise in brain R2* maps and proposes a sampling strategy to mitigate its effects.

This paper develops a deep learning method with uncertainty estimation to detect MRI artifacts in large imbalanced datasets, achieving high accuracy and outperforming previous approaches.

This study compares denoising methods for multi-echo 7T fMRI data, finding that ME-ICA combined with aCompCor and ICA-AROMA is highly effective for noise reduction and signal preservation.

The paper introduces an open-access dataset combining EEG and fMRI data from naturalistic viewing tasks to study brain function and preprocessing methods.

This paper evaluates brain age predictions across repeated time points and finds small within-subject correlations, field strength influence, and model biases.

The paper shows resting-state fMRI signals contain spectral signatures reflecting local hemodynamic response timing variations across brain regions.

MISPEL is a supervised deep learning method for harmonizing multi-scanner neuroimaging data, outperforming existing methods in reducing scanner-related variability.