Markus Barth

Building a high-resolution in vivo minimum deformation average model of the human hippocampus

Publisher: Cold Spring Harbor Laboratory

Date: 06-07-2017

DOI: 10.1101/160176

Abstract: Minimum deformation averaging (MDA) procedures exploit the information contained in inter-in idual variations to generate high-resolution, high-contrast models through iterative model building. However, MDA models built from different image contrasts reside in disparate spaces and their complementary information cannot be utilized easily. The aim of this work was to develop an algorithm for the non-linear alignment of two MDA models with different contrasts to create a high-resolution in vivo model of the human hippoc us with a spatial resolution of 300 μm. A Turbo Spin Echo MDA model covering the hippoc us was contrast matched to a whole-brain MP2RAGE MDA model and aligned using cross-correlation and non-linear transformation. The contrast matching algorithm followed a global voxel location-based approach to estimate the relationship between intensity values of the two models. The performance of the algorithm was evaluated by comparing it to a non-linear registration obtained using mutual information without contrast matching. The complimentary information from both contrasts was then utilized in an automated hippoc al subfield segmentation pipeline. The contrast of the Turbo Spin Echo MDA model could successfully be matched to the MP2RAGE MDA model. Registration using cross correlation provided more accurate alignment of the models compared to a mutual information based approach. The segmentation using ASHS resulted in hippoc al subfield delineations that resembled the tissue boundaries observed in the Turbo Spin Echo MDA model. The developed contrast matching algorithm facilitated the creation of a high-resolution multi-modal in vivo MDA model of the human hippoc us. This model can be used to improve algorithms for hippoc al subfield segmentation and could potentially support the early detection of neurodegenerative diseases.

A study-specific fMRI normalization approach that operates directly on high resolution functional EPI data at 7Tesla

Publisher: Elsevier BV

Date: 10-2014

DOI: 10.1016/J.NEUROIMAGE.2014.06.045

A population‐specific symmetric phase model to automatically analyze susceptibility‐weighted imaging (SWI) phase shifts and phase symmetry in the human brain

Publisher: Wiley

Date: 20-12-2009

DOI: 10.1002/JMRI.22013

Abstract: To create a population-specific symmetric phase model and to evaluate the susceptibility-weighted imaging (SWI) phase in terms of phase shift using different segmentation methods (manual and automatic) and phase shift symmetry, which is expected as a marker for lateralized Parkinson's disease (PD) symptoms. SWI and T(1)-weighted data from 25 PD patients and five healthy controls were acquired on a 3T MRI system. A population-specific, symmetric phase model was developed. Regions of interest (ROIs) were defined manually on the phase model, manually on each in idual data set, and automatically using model-based segmentation (MBS). Manually- and MBS-defined ROIs were compared using kappa values, and left-right phase symmetry was evaluated using correlation analysis. Independent of the analysis method, a phase increase from the anterior to the posterior putamen, and the average phase value relationship substantia nigra > globus pallidus > red nucleus was found. Phase symmetry analysis shows a difference between lateralized and symmetric PD. The symmetric phase model helps to analyze phase data with similar accuracy, but a greatly reduced tracing effort compared to in idual tracing and also allows evaluating left-right phase symmetries.

Measuring the effects of attention to individual fingertips in somatosensory cortex using ultra-high field (7T) fMRI

Publisher: Elsevier BV

Date: 11-2017

DOI: 10.1016/J.NEUROIMAGE.2017.08.014

Abstract: Attention to sensory information has been shown to modulate the neuronal processing of that information. For ex le, visuospatial attention acts by modulating responses at retinotopically appropriate regions of visual cortex (Puckett and DeYoe, 2015 Tootell et al. 1998). Much less, however, is known about the neuronal processing associated with attending to other modalities of sensory information. One reason for this is that visual cortex is relatively large, and therefore easier to access non-invasively in humans using tools such as functional magnetic resonance imaging (fMRI). With high-resolution fMRI, however, it is now possible to access smaller cortical areas such as primary somatosensory cortex (Martuzzi et al., 2014 Sanchez-Panchuelo et al., 2010 Schweisfurth et al. 2014 Schweizer et al. 2008). Here, we combined a novel experimental design and high-resolution fMRI at ultra-high field (7T) to measure the effects of attention to tactile stimulation in primary somatosensory cortex, S1. We find that attention modulates somatotopically appropriate regions of S1, and importantly, that this modulation can be measured at the level of the cortical representation of in idual fingertips.

Super-resolution QSM in little or no additional time for imaging (NATIve) using 2D EPI imaging in 3 orthogonal planes

Publisher: Elsevier BV

Date: 12-2023

DOI: 10.1016/J.NEUROIMAGE.2023.120419

A time‐efficient acquisition protocol for multipurpose diffusion‐weighted microstructural imaging at 7 Tesla

Publisher: Wiley

Date: 12-02-2017

DOI: 10.1002/MRM.26608

Abstract: Several diffusion-weighted MRI techniques have been developed and validated during the past 2 decades. While offering various neuroanatomical inferences, these techniques differ in their proposed optimal acquisition design, preventing clinicians and researchers benefiting from all potential inference methods, particularly when limited time is available. This study reports an optimal design that enables for a time-efficient diffusion-weighted MRI acquisition scheme at 7 Tesla. The primary audience of this article is the typical end user, interested in diffusion-weighted microstructural imaging at 7 Tesla. We tested b-values in the range of 700 to 3000 s/mm The suggested design is a protocol with b-values of 1000 and 2500 s/mm We estimated minimum acquisition requirements that enable diffusion tensor imaging, higher angular resolution diffusion-weighted imaging, neurite orientation dispersion, and density imaging and white matter tract integrity across whole brain with isotropic resolution of 1.8 mm in less than 11 min. Magn Reson Med 78:2170-2184, 2017. © 2017 International Society for Magnetic Resonance in Medicine.

7T GRE-MRI signal compartments are sensitive to dysplastic tissue in focal epilepsy

Publisher: Elsevier BV

Date: 09-2019

DOI: 10.1016/J.MRI.2019.05.011

Abstract: Ultra-high field magnetic resonance imaging data obtained using a multi-echo gradient echo sequence has been shown to contain information on tissue microstructure. Quantitative assessment of water fraction, relaxation time and frequency shift using multi-compartment signal modelling may help improve our understanding of diseases and disorders affecting the human brain. In this study, we explored tissue microstructure information by analysing voxel compartment water fraction and frequency shifts derived from 7 T multi-echo gradient recalled echo MRI data. We aimed to test whether the parameters of a three compartment model could distinguish the normal cortex from the cortex affected by focal cortical dysplasia. We compartmentalised normal and dysplastic cortical regions in patients diagnosed with focal cortical dysplasia. We found the frequency shift parameter of the shortest T

Generic acquisition protocol for quantitative MRI of the spinal cord

Publisher: Springer Science and Business Media LLC

Date: 16-08-2021

DOI: 10.1038/S41596-021-00588-0

Towards Optimising MRI Characterisation of Tissue (TOMCAT) Dataset including all Longitudinal Automatic Segmentation of Hippocampal Subfields (LASHiS) data

Publisher: Elsevier BV

Date: 10-2020

DOI: 10.1016/J.DIB.2020.106043

Author Correction: Open-access quantitative MRI data of the spinal cord and reproducibility across participants, sites and manufacturers

Publisher: Springer Science and Business Media LLC

Date: 10-09-2021

DOI: 10.1038/S41597-021-01026-2

Echo time‐dependent quantitative susceptibility mapping contains information on tissue properties

Publisher: Wiley

Date: 25-05-2016

DOI: 10.1002/MRM.26281

Abstract: Magnetic susceptibility is a physical property of matter that varies depending on chemical composition and abundance of different molecular species. Interest is growing in mapping of magnetic susceptibility in the human brain using magnetic resonance imaging techniques, but the influences affecting the mapped values are not fully understood. We performed quantitative susceptibility mapping on 7 Tesla (T) multiple echo time gradient recalled echo data and evaluated the trend in 10 regions of the human brain. Temporal plots of susceptibility were performed in the caudate, pallidum, putamen, thalamus, insula, red nucleus, substantia nigra, internal capsule, corpus callosum, and fornix. We implemented an existing three compartment signal model and used optimization to fit the experimental result to assess the influences that could be responsible for our findings. The temporal trend in susceptibility is different for different brain regions, and subsegmentation of specific regions suggests that differences are likely to be attributable to variations in tissue structure and composition. Using a signal model, we verified that a nonlinear temporal behavior in experimentally computed susceptibility within imaging voxels may be the result of the heterogeneous composition of tissue properties. Decomposition of voxel constituents into meaningful parameters may lead to informative measures that reflect changes in tissue microstructure. Magn Reson Med 77:1946–1958, 2017. © 2016 International Society for Magnetic Resonance in Medicine

Open-access quantitative MRI data of the spinal cord and reproducibility across participants, sites and manufacturers

Publisher: Springer Science and Business Media LLC

Date: 16-08-2021

DOI: 10.1038/S41597-021-00941-8

Abstract: In a companion paper by Cohen-Adad et al . we introduce the spine generic quantitative MRI protocol that provides valuable metrics for assessing spinal cord macrostructural and microstructural integrity. This protocol was used to acquire a single subject dataset across 19 centers and a multi-subject dataset across 42 centers (for a total of 260 participants), spanning the three main MRI manufacturers: GE, Philips and Siemens. Both datasets are publicly available via git-annex. Data were analysed using the Spinal Cord Toolbox to produce normative values as well as inter/intra-site and inter/intra-manufacturer statistics. Reproducibility for the spine generic protocol was high across sites and manufacturers, with an average inter-site coefficient of variation of less than 5% for all the metrics. Full documentation and results can be found at spine-generic.rtfd.io/ . The datasets and analysis pipeline will help pave the way towards accessible and reproducible quantitative MRI in the spinal cord.

Phase locking of auditory steady state responses is modulated by a predictive sensory context and linked to degree of myelination in the cerebellum

Publisher: Cold Spring Harbor Laboratory

Date: 12-06-2023

DOI: 10.1101/2023.06.08.23291140

Abstract: 40 Hz auditory steady state responses (ASSR) can be evoked by brief auditory clicks delivered at 40 Hz. While the neuropharmacology behind the generation of ASSR is well examined, the link between ASSR and microstructural properties of the brain is unclear. Further, whether the 40 Hz ASSR can be manipulated through processes involving top-down control, such as prediction, is currently unknown. We recorded EEG in 50 neurotypical participants while they engaged in a 40 Hz Auditory steady state paradigm. We manipulated the predictability of tones to test the modulatory effect of prediction on 40 Hz steady state responses. Further, we acquired T1w and T2w structural MRI and used the T1/T2 ratio as a proxy to determine myelination in grey matter. The phase locking of the 40 Hz ASSR was indeed modulated by prediction and this modulation extended to all frequency bands, suggesting prediction violation as a phase resetting mechanism. Interestingly, we found that the prediction violation of the phase locking at 40 Hz (gamma) was associated with the degree of grey matter myelination in the right cerebellum. We demonstrate that prediction violations evoke resetting of oscillatory activity and suggest that the efficiency of this process is promoted by greater cerebellar myelin. Our findings provide a structural-functional relationship for myelin and phase locking of auditory oscillatory activity. These results introduce a setting for looking at the interaction of predictive processes and ASSR in disorders where these processes are impaired such as in psychosis.

DeepQSM - Using Deep Learning to Solve the Dipole Inversion for MRI Susceptibility Mapping

Publisher: Cold Spring Harbor Laboratory

Date: 07-03-2018

DOI: 10.1101/278036

Abstract: Quantitative susceptibility mapping (QSM) aims to extract the magnetic susceptibility of tissue from magnetic resonance imaging (MRI) phase measurements. The mapping of magnetic susceptibility in vivo has gained broad interest in several fields of science and medicine because it yields relevant information on biological tissue properties, predominantly myelin, iron and calcium. Thereby, QSM can also reveal pathological changes of these key components in devastating diseases such as Parkinson’s disease, Multiple Sclerosis, or hepatic iron overload. As QSM requires the solution of an ill-posed field-to-source-inversion, current techniques utilize manual optimization of regularization parameters to balance between smoothing, artifacts and quantification accuracy. We trained a fully convolutional deep neural network - DeepQSM - to invert the magnetic dipole kernel convolution. This network is capable of solving the ill-posed field-to-source inversion on real-world in vivo MRI phase data without the need for manual parameter tuning, which proves that this network has generalized the underlying mathematical principle of the dipole inversion. We demonstrate that DeepQSM’s susceptibility maps enable identification of deep brain substructures that are not visible in MRI phase data and provide information on their respective magnetic tissue properties. We illustrate DeepQSM’s clinical relevance in a patient with multiple sclerosis showing its sensitivity to white matter lesions. In summary, DeepQSM can be used to determine the composition of myelin sheets of nerve fibers in the brain, and to assess quantitative information on iron homeostasis and its dysregulation, and will subsequently contribute to a better understanding of these biological processes in health and disease.

Bayesian population receptive field modeling in human somatosensory cortex

Publisher: Cold Spring Harbor Laboratory

Date: 16-03-2019

DOI: 10.1101/577981

Abstract: Somatosensation is fundamental to our ability to sense our body and interact with the world. Our body is continuously s ling the environment using a variety of receptors tuned to different features, and this information is routed up to primary somatosensory cortex. Strikingly, the spatial organization of the peripheral receptors in the body are well maintained, with the resulting representation of the body in the brain being referred to as the somatosensory homunculus. Recent years have seen considerable advancements in the field of high-resolution fMRI, which have enabled an increasingly detailed examination of the organization and properties of this homunculus. Here we combined advanced imaging techniques at ultra-high field (7T) with a recently developed Bayesian population receptive field (pRF) modeling framework to examine pRF properties in primary somatosensory cortex. In each subject, vibrotactile stimulation of the fingertips (i.e., the peripheral mechanoreceptors) modulated the fMRI response along the post-central gyrus and these signals were used to estimate pRFs. We found the pRF center location estimates to be in accord with previous work as well as evidence of other properties in line with the underlying neurobiology. Specifically, as expected from the known properties of cortical magnification, we find a larger representation of the index finger compared to the other stimulated digits (middle, index, little). We also show evidence that the little finger is marked by the largest pRF sizes. The ability to estimate somatosensory pRFs in humans provides an unprecedented opportunity to examine the neural mechanisms underlying somatosensation and is critical for studying how the brain, body, and environment interact to inform perception and action.

SHARQnet - Sophisticated Harmonic Artifact Reduction in Quantitative Susceptibility Mapping using a Deep Convolutional Neural Network

Publisher: Cold Spring Harbor Laboratory

Date: 17-01-2019

DOI: 10.1101/522151

Abstract: Quantitative susceptibility mapping (QSM) reveals pathological changes in widespread diseases such as Parkinson’s disease, Multiple Sclerosis, or hepatic iron overload. QSM requires multiple processing steps after the acquisition of magnetic resonance imaging (MRI) phase measurements such as unwrapping, background field removal and the solution of an ill-posed field-to-source-inversion. Current techniques utilize iterative optimization procedures to solve the inversion and background field correction, which are computationally expensive and lead to suboptimal or over-regularized solutions requiring a careful choice of parameters that make a clinical application of QSM challenging. We have previously demonstrated that a deep convolutional neural network can invert the magnetic dipole kernel with a very efficient feed forward multiplication not requiring iterative optimization or the choice of regularization parameters. In this work, we extended this approach to remove background fields in QSM. The prototype method, called SHARQnet, was trained on simulated background fields and tested on 3T and 7T brain datasets. We show that SHARQnet outperforms current background field removal procedures and generalizes to a wide range of input data without requiring any parameter adjustments. In summary, we demonstrate that the solution of ill-posed problems in QSM can be achieved by learning the underlying physics causing the artifacts and removing them in an efficient and reliable manner and thereby will help to bring QSM towards clinical applications.

Assessment of microstructural signal compartments across the corpus callosum using multi-echo gradient recalled echo at 7 T

Publisher: Elsevier BV

Date: 11-2018

DOI: 10.1016/J.NEUROIMAGE.2017.11.029

Abstract: Quantitative assessment of tissue microstructure is important in studying human brain diseases and disorders in which white matter is implicated, as it has been linked to demyelination, re-myelination, and axonal damage in clinical conditions. Ultra-high field magnetic resonance imaging data obtained using a multi-echo gradient echo sequence has been shown to contain information on myelin, axonal and extracellular compartments in white matter. In this study, we aimed to assess the sensitivity of a three-compartment model to estimate the variation of corresponding compartment parameters (water fraction, relaxation time and frequency shift) of the corpus callosum sub-regions, which are known to have different tissue structure. Additionally, we computed the g-ratio using myelin and axonal water fractions and performed a voxel-by-voxel analysis in the corpus callosum. Based on data acquired for ten participants, we show that the myelin compartment water fraction and T

Patient with ALS with a novel TBK1 mutation, widespread brain involvement, behaviour changes and metabolic dysfunction

Publisher: BMJ

Date: 08-09-2018

DOI: 10.1136/JNNP-2018-318823

HumanBrainAtlas: an in vivo MRI dataset for detailed segmentations

Publisher: Cold Spring Harbor Laboratory

Date: 18-10-2022

DOI: 10.1101/2022.10.16.511844

Abstract: We introduce HumanBrainAtlas, an initiative to construct a highly detailed, open-access atlas of the living human brain that combines high-resolution in vivo MR imaging and detailed segmentations previously possible only in histological preparations. Here, we present and evaluate the first step of this initiative: a comprehensive dataset of two healthy male volunteers reconstructed to a 0.25 mm 3 isotropic resolution for T1w, T2w and DWI contrasts. Multiple high-resolution acquisitions were collected for each contrast and each participant, followed by averaging using symmetric group-wise normalisation (Advanced Normalisation Tools). The resulting image quality permits structural parcellations rivalling histology-based atlases, while maintaining the advantages of in vivo MRI. For ex le, components of the thalamus, hypothalamus, and hippoc us - difficult or often impossible to identify using standard MRI protocols, can be identified within the present data. Our data are virtually distortion free, fully 3D, and compatible with existing in vivo Neuroimaging analysis tools. The dataset is suitable for teaching and is publicly available via our website ( www.hba.neura.edu.au ), which also provides data processing scripts. Instead of focusing on coordinates in an averaged brain space, our approach focuses on providing an ex le segmentation at great detail in the high quality in idual brain, this serves as an illustration on what features contrasts and relations can be used to interpret MRI datasets, in research, clinical and education settings.

Author Correction: Open-access quantitative MRI data of the spinal cord and reproducibility across participants, sites and manufacturers

Publisher: Springer Science and Business Media LLC

Date: 23-09-2021

DOI: 10.1038/S41597-021-01044-0

SHARQnet - Sophisticated Harmonic Artifact Reduction in Quantitative Susceptibility Mapping using a Deep Convolutional Neural Network

Publisher: Elsevier BV

Date: 05-2019

DOI: 10.1016/J.ZEMEDI.2019.01.001

Abstract: Quantitative susceptibility mapping (QSM) reveals pathological changes in widespread diseases such as Parkinson's disease, Multiple Sclerosis, or hepatic iron overload. QSM requires multiple processing steps after the acquisition of magnetic resonance imaging (MRI) phase measurements such as unwrapping, background field removal and the solution of an ill-posed field-to-source-inversion. Current techniques utilize iterative optimization procedures to solve the inversion and background field correction, which are computationally expensive and lead to suboptimal or over-regularized solutions requiring a careful choice of parameters that make a clinical application of QSM challenging. We have previously demonstrated that a deep convolutional neural network can invert the magnetic dipole kernel with a very efficient feed forward multiplication not requiring iterative optimization or the choice of regularization parameters. In this work, we extended this approach to remove background fields in QSM. The prototype method, called SHARQnet, was trained on simulated background fields and tested on 3T and 7T brain datasets. We show that SHARQnet outperforms current background field removal procedures and generalizes to a wide range of input data without requiring any parameter adjustments. In summary, we demonstrate that the solution of ill-posed problems in QSM can be achieved by learning the underlying physics causing the artifacts and removing them in an efficient and reliable manner and thereby will help to bring QSM towards clinical applications.

Non-linear realignment improves hippocampus subfield segmentation reliability

Publisher: Elsevier BV

Date: 12-2019

DOI: 10.1016/J.NEUROIMAGE.2019.116206

Abstract: Participant movement can deleteriously affect MR image quality. Further, for the visualization and segmentation of small anatomical structures, there is a need to improve image quality, specifically signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), by acquiring multiple anatomical scans consecutively. We aimed to ameliorate movement artefacts and increase SNR in a high-resolution turbo spin-echo (TSE) sequence acquired thrice using non-linear realignment in order to improve segmentation consistency of the hippoc us subfields. We assessed the method in 29 young healthy participants, 11 Motor Neuron Disease patients, and 11 age matched controls at 7T, and 24 healthy adolescents at 3T. Results show improved image segmentation of the hippoc us subfields when comparing template-based segmentations with in idual segmentations with Dice overlaps N = 75 ps < 0.001 (Friedman's test) and higher sharpness ps < 0.001 in non-linearly realigned scans as compared to linearly, and arithmetically averaged scans.

The challenge of bias‐free coil combination for quantitative susceptibility mapping at ultra‐high field

Publisher: Wiley

Date: 03-2017

DOI: 10.1002/MRM.26644

Abstract: Quantitative susceptibility mapping is a technique to estimate the magnetic property of tissue with particularly high sensitivity at ultra-high field. However, a key challenge at ultra-high field is the combination of phase data acquired using phased array receive coils. Several methods for combining phase data have been proposed, but the influence of coil combination choices on susceptibility quantitation has not been studied systematically. We combined phase data using COMPOSER (COMbining Phase data using a Short Echo-time Reference scan) and a reference-free channel-by-channel method. We investigated the effect of the chosen combination method on susceptibility results in a group of 28 participants at 7 Tesla. Our results show that reference scans can bias susceptibility values. Although the proposed reference-free channel-by-channel method cannot remove transmit field phase, it shows comparable results to the COMPOSER method in which a high-resolution ultrashort echo-time reference scan was used. We conclude that ultrashort echo-time reference scans reduce quantitation bias and remove the transmit field phase when using COMPOSER to combine phase data, and not combining the phase data before susceptibility processing avoids this bias, resulting in comparable results. Magn Reson Med 79:97-107, 2018. © 2017 InternationalSociety for Magnetic Resonance in Medicine.

DeepQSM - using deep learning to solve the dipole inversion for quantitative susceptibility mapping

Publisher: Elsevier BV

Date: 07-2019

DOI: 10.1016/J.NEUROIMAGE.2019.03.060

Abstract: Quantitative susceptibility mapping (QSM) is based on magnetic resonance imaging (MRI) phase measurements and has gained broad interest because it yields relevant information on biological tissue properties, predominantly myelin, iron and calcium in vivo. Thereby, QSM can also reveal pathological changes of these key components in widespread diseases such as Parkinson's disease, Multiple Sclerosis, or hepatic iron overload. While the ill-posed field-to-source-inversion problem underlying QSM is conventionally assessed by the means of regularization techniques, we trained a fully convolutional deep neural network - DeepQSM - to directly invert the magnetic dipole kernel convolution. DeepQSM learned the physical forward problem using purely synthetic data and is capable of solving the ill-posed field-to-source inversion on in vivo MRI phase data. The magnetic susceptibility maps reconstructed by DeepQSM enable identification of deep brain substructures and provide information on their respective magnetic tissue properties. In summary, DeepQSM can invert the magnetic dipole kernel convolution and delivers robust solutions to this ill-posed problem.

ECG Triggering in Ultra-High Field Cardiovascular MRI

Publisher: MDPI AG

Date: 09-2016

DOI: 10.18383/J.TOM.2016.00193

Abstract: Cardiac magnetic resonance imaging at ultra-high field (B0 ≥ 7 T) potentially provides improved resolution and new opportunities for tissue characterization. Although an accurate synchronization of the acquisition to the cardiac cycle is essential, electrocardiogram (ECG) triggering at ultra-high field can be significantly impacted by the magnetohydrodynamic (MHD) effect. Blood flow within a static magnetic field induces a voltage, which superimposes the ECG and often affects the recognition of the R-wave. The MHD effect scales with B0 and is particularly pronounced at ultra-high field creating triggering-related image artifacts. Here, we investigated the performance of a conventional 3-lead ECG trigger device and a state-of-the-art trigger algorithm for cardiac ECG synchronization at 7 T. We show that by appropriate subject preparation and by including a learning phase for the R-wave detection outside of the magnetic field, reliable ECG triggering is feasible in healthy subjects at 7 T without additional equipment. Ultra-high field cardiac imaging was performed with the ECG signal and the trigger events recorded in 8 healthy subjects. Despite severe ECG signal distortions, synchronized imaging was successfully performed. Recorded ECG signals, vectorcardiograms, and large consistency in trigger event spacing indicate high accuracy for R-wave detection.

Filtered deconvolution of a simulated and an in vivo phase model of the human brain

Publisher: Wiley

Date: 22-07-2010

DOI: 10.1002/JMRI.22246

Abstract: To remove spatial patterns in gradient echo phase images which are caused by susceptibility differences between different tissue types using filtered deconvolution and to evaluate deconvolution effects. A realistic simulated susceptibility map of the human brain was built and used to evaluate the effects of filtered deconvolution. The simulated susceptibility map was convolved with a filter kernel representing a magnetic dipole resulting in a simulated phase map. The artificial phase map was superimposed with different noise levels and deconvolved using different deconvolution kernels. The resulting contrast-to-noise ratios between white and gray matter of the deconvolved data provide an estimate for an optimal deconvolution kernel for a given noise level. These results were used to deconvolve an in vivo phase model representing the average of 30 phase data sets and also in idual phase data acquired at 7 Tesla. The deconvolved phase model shows a better anatomical agreement with the corresponding magnitude than the original phase model (5% higher kappa coefficient). Visual inspection of the deconvolved in idual phase shows a more consistent delineation of blood vessels. Filtered deconvolution of SWI phase is possible when an appropriate filter kernel is used. This helps to improve region of interest definition as unrealistic phase patterns are removed.

Non-Linear Realignment Improves Hippocampus Subfield Segmentation

Publisher: Cold Spring Harbor Laboratory

Date: 05-04-2019

DOI: 10.1101/597856

Abstract: Participant movement can deleteriously affect MR image quality. Further, for the visualization and segmentation of small anatomical structures, there is a need to improve image quality, specifically signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), by acquiring multiple anatomical scans consecutively. We aimed to ameliorate movement artefacts and increase SNR in a high-resolution turbo spin-echo (TSE) sequence acquired thrice using non-linear realignment in order to improve segmentation consistency of the hippoc us subfields. We assessed the method in young healthy participants, Motor Neurone Disease patients, and age matched controls. Results show improved image segmentation of the hippoc us subfields when comparing template-based segmentations with in idual segmentations with Dice overlaps N=51 ps 0.001 (Friedman’s test) and higher sharpness ps 0.001 in non-linearly realigned scans as compared to linearly, and arithmetically averaged scans.

Longitudinal Automatic Segmentation of Hippocampal Subfields (LASHiS) using Multi-Contrast MRI

Publisher: Cold Spring Harbor Laboratory

Date: 08-09-2019

DOI: 10.1101/759217

Abstract: The volumetric and morphometric examination of hippoc us formation subfields in a longitudinal manner using in vivo MRI could lead to more sensitive biomarkers for neuropsychiatric disorders and diseases including Alzheimer’s disease, as the anatomical subregions are functionally specialised. Longitudinal processing allows for increased sensitivity due to reduced confounds of inter-subject variability and higher effect-sensitivity than cross-sectional designs. We examined the performance of a new longitudinal pipeline (Longitudinal Automatic Segmentation of Hippoc us Subfields [LASHiS]) against three freely available, published approaches. LASHiS automatically segments hippoc us formation subfields by propagating labels from cross-sectionally labelled time point scans using joint-label fusion to a non-linearly realigned ‘single subject template’, where image segmentation occurs free of bias to any in idual time point. Our pipeline measures tissue characteristics available in in vivo high-resolution MRI scans, at both clinical (3 Tesla) and ultra-high field strength (7 Tesla) and differs from previous longitudinal segmentation pipelines in that it leverages multi-contrast information in the segmentation process. LASHiS produces robust and reliable automatic multi-contrast segmentations of hippoc us formation subfields, as measured by higher volume similarity coefficients and Dice coefficients for test-retest reliability and robust longitudinal Bayesian Linear Mixed Effects results at 7 T, while showing sound results at 3 T. All code for this project including the automatic pipeline is available at github.com/CAIsr/LASHiS

7-Tesla Functional Cardiovascular MR Using Vectorcardiographic Triggering – overcoming the magnetohydrodynamic effect

Publisher: Cold Spring Harbor Laboratory

Date: 14-05-2021

DOI: 10.1101/2021.05.10.21256949

Abstract: Ultra-high-field (B 0 ≥ 7 Tesla) cardiovascular magnetic resonance (CMR) offers increased resolution. However, ECG gating is impacted by the magneto-hydrodynamic (MHD) effect distorting the ECG trace. We explored the technical feasibility of a 7T MR scanner using ECG trigger learning algorithm to quantitatively assess cardiac volumes and vascular flow. 7T scans performed on 10 healthy volunteers on a whole-body research MRI (Siemens Healthcare, Erlangen, Germany) with 8 channel Tx/32 channel Rx cardiac coil (MRI Tools GmbH, Berlin, Germany). Vectorcardiogram ECG was performed using a learning phase outside of the magnetic field, with a trigger algorithm overcoming severe ECG signal distortions. Vectorcardiograms were quantitatively analyzed for false negative and false positive events. Cine CMR was performed after 3 rd -order B 0 shimming using a high-resolution breath-held ECG-retro-gated segmented spoiled gradient echo, and 2D phase contrast flow imaging. Artefacts were assessed using a semi-quantitative scale. 7T CMR scans were acquired in all patients (100%) using the VCG learning method. 3,142 R-waves were quantitatively analyzed, yielding sensitivity 97.6%, specificity 98.7%. Mean image quality score was 0.9, sufficient to quantitate both cardiac volumes, ejection fraction (EF), aortic and pulmonary blood flow. Mean LVEF was 56.4%, RVEF 51.4%. Reliable cardiac ECG triggering is feasible in healthy volunteers at 7T utilizing a state-of-the-art 3-lead trigger device despite signal distortion from the MHD effect. This provides sufficient image quality for quantitative analysis. Other ultra-high-field imaging applications such as human brain functional MRI with physiologic noise correction may benefit from this method of ECG triggering. Ultra-high field 7 Tesla cardiac MRI is challenging due to the impact of the magneto-hydrodynamic (MHD) effect causing severe distortions in the ECG trace. Using VCG with a learning phase outside the ultra-high field magnet, the R waves can be adequately detected to perform high quality Cardiac MRI scans, overcoming signal distortion from the MHD effect.

Using multi-echo simultaneous multi-slice (SMS) EPI to improve functional MRI of the subcortical nuclei of the basal ganglia at ultra-high field (7T)

Publisher: Elsevier BV

Date: 05-2018

DOI: 10.1016/J.NEUROIMAGE.2017.12.005

Abstract: The nuclei of the basal ganglia pose a special problem for functional MRI, especially at ultra-high field, because T2* variations between different regions result in suboptimal BOLD sensitivity when using gradient-echo echo-planar imaging (EPI). Specifically, the iron-rich lentiform nucleus of the basal ganglia, including the putamen and globus pallidus, suffers from substantial signal loss when imaging is performed using conventional single-echo EPI with echo times optimized for the cortex. Multi-echo EPI acquires several echoes at different echo times for every imaging slice, allowing images to be reconstructed with a weighting of echo times that is optimized in idually for each voxel according to the underlying tissue or T2* properties. Here we show that multi-echo simultaneous multi-slice (SMS) EPI can improve functional activation of iron-rich subcortical regions while maintaining sensitivity within cortical areas. Functional imaging during a motor task known to elicit strong activations in the cortex and the subcortex (basal ganglia) was performed to compare the performance of multi-echo SMS EPI to single-echo SMS EPI. Notably within both the caudate nucleus and putamen of the basal ganglia, multi-echo SMS EPI yielded higher tSNR (an average 84% increase) and CNR (an average 58% increase), an approximate 3-fold increase in supra-threshold voxels, and higher t-values (an average 39% increase). The degree of improvement in the group level t-statistics was negatively correlated to the underlying T2* of the voxels, such that the shorter the T2*, as in the iron-rich nuclei of the basal ganglia, the higher the improvement of t-values in the activated region.

Serial correlations in single-subject fMRI with sub-second TR

Publisher: Elsevier BV

Date: 02-2018

DOI: 10.1016/J.NEUROIMAGE.2017.10.043

Abstract: When performing statistical analysis of single-subject fMRI data, serial correlations need to be taken into account to allow for valid inference. Otherwise, the variability in the parameter estimates might be under-estimated resulting in increased false-positive rates. Serial correlations in fMRI data are commonly characterized in terms of a first-order autoregressive (AR) process and then removed via pre-whitening. The required noise model for the pre-whitening depends on a number of parameters, particularly the repetition time (TR). Here we investigate how the sub-second temporal resolution provided by simultaneous multislice (SMS) imaging changes the noise structure in fMRI time series. We fit a higher-order AR model and then estimate the optimal AR model order for a sequence with a TR of less than 600 ms providing whole brain coverage. We show that physiological noise modelling successfully reduces the required AR model order, but remaining serial correlations necessitate an advanced noise model. We conclude that commonly used noise models, such as the AR(1) model, are inadequate for modelling serial correlations in fMRI using sub-second TRs. Rather, physiological noise modelling in combination with advanced pre-whitening schemes enable valid inference in single-subject analysis using fast fMRI sequences.

Combining phase images from array coils using a short echo time reference scan (COMPOSER)

Publisher: Wiley

Date: 29-12-2015

DOI: 10.1002/MRM.26093

Open-Source Hypothalamic-ForniX (OSHy-X) Atlases and Segmentation Tool for 3T and 7T

Publisher: The Open Journal

Date: 05-08-2022

DOI: 10.21105/JOSS.04368

Improving FLAIR SAR efficiency at 7T by adaptive tailoring of adiabatic pulse power through deep learning estimation

Publisher: Wiley

Date: 23-11-2020

DOI: 10.1002/MRM.28590

Abstract: The purpose of this study is to demonstrate a method for specific absorption rate (SAR) reduction for 2D T 2 ‐FLAIR MRI sequences at 7 T by predicting the required adiabatic radiofrequency (RF) pulse power and scaling the RF litude in a slice‐wise fashion. We used a time‐res led frequency‐offset corrected inversion (TR‐FOCI) adiabatic pulse for spin inversion in a T 2 ‐FLAIR sequence to improve homogeneity and calculated the pulse power required for adiabaticity slice‐by‐slice to minimize the SAR. Drawing on the implicit inhomogeneity in a standard localizer scan, we acquired 3D AutoAlign localizers and SA2RAGE maps in 28 volunteers. Then, we trained a convolutional neural network (CNN) to estimate the profile from the localizers and calculated pulse scale factors for each slice. We assessed the predicted profiles and the effect of scaled pulse litudes on the FLAIR inversion efficiency in oblique transverse, sagittal, and coronal orientations. The predicted litude maps matched the measured ones with a mean difference of 9.5% across all slices and participants. The slice‐by‐slice scaling of the TR‐FOCI inversion pulse was most effective in oblique transverse orientation and resulted in a 1 min and 30 s reduction in SAR induced delay time while delivering identical image quality. We propose a SAR reduction technique based on the estimation of profiles from standard localizer scans using a CNN and show that scaling the inversion pulse power slice‐by‐slice for FLAIR sequences at 7T reduces SAR and scan time without compromising image quality.

New acquisition techniques and their prospects for the achievable resolution of fMRI

Publisher: Elsevier BV

Date: 12-2020

DOI: 10.1016/J.PNEUROBIO.2020.101936

Improved dynamic distortion correction for fMRI using single‐echo EPI and a readout‐reversed first image (REFILL)

Publisher: Wiley

Date: 07-08-2023

DOI: 10.1002/HBM.26440

Abstract: The boundaries between tissues with different magnetic susceptibilities generate inhomogeneities in the main magnetic field which change over time due to motion, respiration and system instabilities. The dynamically changing field can be measured from the phase of the fMRI data and corrected. However, methods for doing so need multi‐echo data, time‐consuming reference scans and/or involve error‐prone processing steps, such as phase unwrapping, which are difficult to implement robustly on the MRI host. The improved dynamic distortion correction method we propose is based on the phase of the single‐echo EPI data acquired for fMRI, phase offsets calculated from a triple‐echo, bipolar reference scan of circa 3–10 s duration using a method which avoids the need for phase unwrapping and an additional correction derived from one EPI volume in which the readout direction is reversed. This Reverse‐Encoded First Image and Low resoLution reference scan (REFILL) approach is shown to accurately measure B 0 as it changes due to shim, motion and respiration, even with large dynamic changes to the field at 7 T, where it led to a 20% increase in time‐series signal to noise ratio compared to data corrected with the classic static approach. fMRI results from REFILL‐corrected data were free of stimulus‐correlated distortion artefacts seen when data were corrected with static field mapping. The method is insensitive to shim changes and eddy current differences between the reference scan and the fMRI time series, and employs calculation steps that are simple and robust, allowing most data processing to be performed in real time on the scanner image reconstruction computer. These improvements make it feasible to routinely perform dynamic distortion correction in fMRI.

Selective channel combination of MRI signal phase

Publisher: Wiley

Date: 24-11-2015

DOI: 10.1002/MRM.26057

Abstract: Signal magnitude can robustly be combined using the sum-of-squares approach. Methods have been developed to combine complex images. However, techniques based only on signal phase have not been developed and evaluated. We performed simulations to demonstrate the effect of noise on coil combination. 32-channel 7 Tesla human gradient echo MRI brain data were collected. We combined phase images based on phase noise leading to spatially selective and coil selective combination of phase images. We compared our selective combination approach to optimal noise distribution and adaptive combination methods. We found that selective combination of signal phases leads to improved phase signal-to-noise ratio. Furthermore, a phase shift can be present in combined phase images introduced by the method used to combine multiple channel phases. Mapping of signal phase from ultra-high field MRI data undoubtedly provides a wealth of information about the ageing brain and the effects of neurodegenerative disorders. Measurement of signal phase is essential in frequency shift mapping and in quantitative susceptibility mapping. The method used to combine signal phase should be informed by an understanding of the noise distribution in signal phase at the in idual channel level. Magn Reson Med 76:1469-1477, 2016. © 2015 International Society for Magnetic Resonance in Medicine.

Lower hypothalamic volume with lower body mass index is associated with shorter survival in patients with amyotrophic lateral sclerosis

Publisher: Wiley

Date: 08-11-2023

DOI: 10.1111/ENE.15589

Abstract: Weight loss in patients with amyotrophic lateral sclerosis (ALS) is associated with faster disease progression and shorter survival. Decreased hypothalamic volume is proposed to contribute to weight loss due to loss of appetite and/or hypermetabolism. We aimed to investigate the relationship between hypothalamic volume and body mass index (BMI) in ALS and Alzheimer's disease (AD), and the associations of hypothalamic volume with weight loss, appetite, metabolism and survival in patients with ALS. We compared hypothalamic volumes from magnetic resonance imaging scans with BMI for patients with ALS ( n = 42), patients with AD ( n = 167) and non‐neurodegenerative disease controls ( n = 527). Hypothalamic volumes from patients with ALS were correlated with measures of appetite and metabolism, and change in anthropomorphic measures and disease outcomes. Lower hypothalamic volume was associated with lower and higher BMI in ALS (quadratic association probability of direction = 0.96). This was not observed in AD patients or controls. Hypothalamic volume was not associated with loss of appetite ( p = 0.58) or hypermetabolism ( p = 0.49). Patients with lower BMI and lower hypothalamic volume tended to lose weight ( p = 0.08) and fat mass ( p = 0.06) over the course of their disease, and presented with an increased risk of earlier death (hazard ratio [HR] 3.16, p = 0.03). Lower hypothalamic volume alone trended for greater risk of earlier death (HR 2.61, p = 0.07). These observations suggest that lower hypothalamic volume in ALS contributes to positive and negative energy balance, and is not universally associated with loss of appetite or hypermetabolism. Critically, lower hypothalamic volume with lower BMI was associated with weight loss and earlier death.

Pulsed arterial spin labelling at ultra-high field with a B 1 + -optimised adiabatic labelling pulse

Publisher: Springer Science and Business Media LLC

Date: 15-04-2016

DOI: 10.1007/S10334-016-0555-2

Abstract: Arterial spin labelling (ASL) techniques benefit from the increased signal-to-noise ratio and the longer T 1 relaxation times available at ultra-high field. Previous pulsed ASL studies at 7 T concentrated on the superior regions of the brain because of the larger transmit radiofrequency inhomogeneity experienced at ultra-high field that hinders an adequate inversion of the blood bolus when labelling in the neck. Recently, researchers have proposed to overcome this problem with either the use of dielectric pads, through dedicated transmit labelling coils, or special adiabatic inversion pulses. We investigate the performance of an optimised time-res led frequency-offset corrected inversion (TR-FOCI) pulse designed to cause inversion at much lower peak B 1 (+) . In combination with a PICORE labelling, the perfusion signal obtained with this pulse is compared against that obtained with a FOCI pulse, with and without dielectric pads. Mean grey matter perfusion with the TR-FOCI was 52.5 ± 10.3 mL/100 g/min, being significantly higher than the 34.6 ± 2.6 mL/100 g/min obtained with the FOCI pulse. No significant effect of the dielectric pads was observed. The usage of the B 1 (+) -optimised TR-FOCI pulse results in a significantly higher perfusion signal. PICORE-ASL is feasible at ultra-high field with no changes to operating conditions.

ARC Training Centre For Innovation In Biomedical Imaging Technology

Start Date: 2017

End Date: 2021

Funder: Australian Research Council

ARC Future Fellowships - Grant ID: FT140100865

Start Date: 03-2015

End Date: 12-2022

Amount: $870,552.00

Funder: Australian Research Council

Discovery Projects - Grant ID: DP200103386

Start Date: 2020

End Date: 12-2024

Amount: $512,774.00

Funder: Australian Research Council

Linkage Projects - Grant ID: LP200301393

Start Date: 02-2022

End Date: 02-2025

Amount: $380,115.00

Funder: Australian Research Council

Industrial Transformation Training Centres - Grant ID: IC170100035

Start Date: 12-2017

End Date: 06-2024

Amount: $4,743,710.00

Funder: Australian Research Council