Quantitative susceptibility mapping (QSM) has gained broad interest in the field by extracting bulk tissue magnetic susceptibility, predominantly determined by myelin, iron and calcium from magnetic ...resonance imaging (MRI) phase measurements in vivo. Thereby, QSM can reveal pathological changes of these key components in a variety of diseases. QSM requires multiple processing steps such as phase unwrapping, background field removal and field‐to‐source inversion. Current state‐of‐the‐art techniques utilize iterative optimization procedures to solve the inversion and background field correction, which are computationally expensive and require a careful choice of regularization parameters. With the recent success of deep learning using convolutional neural networks for solving ill‐posed reconstruction problems, the QSM community also adapted these techniques and demonstrated that the QSM processing steps can be solved by efficient feed forward multiplications not requiring either iterative optimization or the choice of regularization parameters. Here, we review the current status of deep learning‐based approaches for processing QSM, highlighting limitations and potential pitfalls, and discuss the future directions the field may take to exploit the latest advances in deep learning for QSM.
We review the current status of deep learning‐based approaches for processing QSM, highlighting limitations and potential pitfalls, and discuss the future directions the field may take to exploit the latest advances in deep learning for QSM.
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.
Display omitted
•A Toolbox to integrate preprocessing of physiological data and fMRI noise modeling.•Robust preprocessing via iterative peak detection, shown for noisy data and patients.•Flexible ...support of peripheral data formats and noise models (RETROICOR, RVHRCOR).•Fully automated noise correction and performance assessment for group studies.•Integration in fMRI pre-processing pipelines as SPM Toolbox (Batch Editor GUI).
Physiological noise is one of the major confounds for fMRI. A common class of correction methods model noise from peripheral measures, such as ECGs or pneumatic belts. However, physiological noise correction has not emerged as a standard preprocessing step for fMRI data yet due to: (1) the varying data quality of physiological recordings, (2) non-standardized peripheral data formats and (3) the lack of full automatization of processing and modeling physiology, required for large-cohort studies.
We introduce the PhysIO Toolbox for preprocessing of physiological recordings and model-based noise correction. It implements a variety of noise models, such as RETROICOR, respiratory volume per time and heart rate variability responses (RVT/HRV). The toolbox covers all intermediate steps − from flexible read-in of data formats to GLM regressor/contrast creation − without any manual intervention.
We demonstrate the workflow of the toolbox and its functionality for datasets from different vendors, recording devices, field strengths and subject populations. Automatization of physiological noise correction and performance evaluation are reported in a group study (N=35).
The PhysIO Toolbox reproduces physiological noise patterns and correction efficacy of previously implemented noise models. It increases modeling robustness by outperforming vendor-provided peak detection methods for physiological cycles. Finally, the toolbox offers an integrated framework with full automatization, including performance monitoring, and flexibility with respect to the input data.
Through its platform-independent Matlab implementation, open-source distribution, and modular structure, the PhysIO Toolbox renders physiological noise correction an accessible preprocessing step for fMRI data.
Whether it be in a single neuron or a more complex biological system like the human brain, form and function are often directly related. The functional organization of human visual cortex, for ...instance, is tightly coupled with the underlying anatomy with cortical shape having been shown to be a useful predictor of the retinotopic organization in early visual cortex. Although the current state-of-the-art in predicting retinotopic maps is able to account for gross individual differences, such models are unable to account for any idiosyncratic differences in the structure-function relationship from anatomical information alone due to their initial assumption of a template. Here we developed a geometric deep learning model capable of exploiting the actual structure of the cortex to learn the complex relationship between brain function and anatomy in human visual cortex such that more realistic and idiosyncratic maps could be predicted. We show that our neural network was not only able to predict the functional organization throughout the visual cortical hierarchy, but that it was also able to predict nuanced variations across individuals. Although we demonstrate its utility for modeling the relationship between structure and function in human visual cortex, our approach is flexible and well-suited for a range of other applications involving data structured in non-Euclidean spaces.
Display omitted
The volumetric and morphometric examination of hippocampus 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 Hippocampus Subfields LASHiS) against three freely available, published approaches. LASHiS automatically segments hippocampus 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 individual time point. Our pipeline measures tissue characteristics available in in vivo high-resolution MRI scans, at both clinical (3 T) and ultra-high field strength (7 T) 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 hippocampus 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 https://github.com/CAIsr/LASHiS.
Display omitted
•LASHiS is an automatic pipeline for segmenting hippocampus subfields.•We utilize multi-contrast information to segment subfields and incorporate joint-label fusion for labelling.•LASHiS out-performs other available methods for automatic hippocampal subfield segmentation in test-retest reliability.•We test LASHiS on both 7 T and 3 T MRI data.
Visual field maps in human early extrastriate areas (V2 and V3) are traditionally thought to form mirror-image representations which surround the primary visual cortex (V1). According to this scheme, ...V2 and V3 form nearly symmetrical halves with respect to the calcarine sulcus, with the dorsal halves representing lower contralateral quadrants, and the ventral halves representing upper contralateral quadrants. This arrangement is considered to be consistent across individuals, and thus predictable with reasonable accuracy using templates. However, data that deviate from this expected pattern have been observed, but mainly treated as artifactual. Here, we systematically investigate individual variability in the visual field maps of human early visual cortex using the 7T Human Connectome Project (HCP) retinotopy dataset. Our results demonstrate substantial and principled inter-individual variability. Visual field representation in the dorsal portions of V2 and V3 was more variable than in their ventral counterparts, including substantial departures from the expected mirror-symmetrical patterns. In addition, left hemisphere retinotopic maps were more variable than those in the right hemisphere. Surprisingly, only one-third of individuals had maps that conformed to the expected pattern in the left hemisphere. Visual field sign analysis further revealed that in many individuals the area conventionally identified as dorsal V3 shows a discontinuity in the mirror-image representation of the retina, associated with a Y-shaped lower vertical representation. Our findings challenge the current view that inter-individual variability in early extrastriate cortex is negligible, and that the dorsal portions of V2 and V3 are roughly mirror images of their ventral counterparts.
•Super-resolution isotropic QSM from three orthogonal 2D SMS EPI acquisitions.•NATIve is 25–50 times faster than 3D GRE.•NATIve QSM values agreed well with 3D GRE values in volunteers and MS ...patients.•The same MS lesions were depicted by 3D GRE and NATIve QSMs.•NATIve is extremely insensitive to motion and achieves high SNR.
Quantitative Susceptibility Mapping has the potential to provide additional insights into neurological diseases but is typically based on a quite long (5–10 min) 3D gradient-echo scan which is highly sensitive to motion. We propose an ultra-fast acquisition based on three orthogonal (sagittal, coronal and axial) 2D simultaneous multi-slice EPI scans with 1 mm in-plane resolution and 3 mm thick slices. Images in each orientation are corrected for susceptibility-related distortions and co-registered with an iterative non-linear Minimum Deformation Averaging (Volgenmodel) approach to generate a high SNR, super-resolution data set with an isotropic resolution of close to 1 mm. The net acquisition time is 3 times the volume acquisition time of EPI or about 12 s, but the three volumes could also replace “dummy scans” in fMRI, making it feasible to acquire QSM in little or No Additional Time for Imaging (NATIve). NATIve QSM values agreed well with reference 3D GRE QSM in the basal ganglia in healthy subjects. In patients with multiple sclerosis, there was also a good agreement between the susceptibility values within lesions and control ROIs and all lesions which could be seen on 3D GRE QSMs could also be visualized on NATIve QSMs. The approach is faster than conventional 3D GRE by a factor of 25–50 and faster than 3D EPI by a factor of 3–5. As a 2D technique, NATIve QSM was shown to be much more robust to motion than the 3D GRE and 3D EPI, opening up the possibility of studying neurological diseases involving iron accumulation and demyelination in patients who find it difficult to lie still for long enough to acquire QSM data with conventional methods.
Display omitted
Quantitative assessment of tissue microstructure is important in studying human brain diseases and disorders. Ultra-high field magnetic resonance imaging (MRI) data obtained using a multi-echo ...gradient echo sequence have been shown to contain information on myelin, axonal, and extracellular compartments in tissue. Quantitative assessment of water fraction, relaxation time (T
*), and frequency shift using multi-compartment models has been shown to be useful in studying white matter properties
specific tissue parameters. It remains unclear how tissue parameters vary with model selection based on 7T multiple echo time gradient-recalled echo (GRE) MRI data. We applied existing signal compartment models to the corpus callosum and investigated whether a three-compartment model can be reduced to two compartments and still resolve white matter parameters i.e., myelin water fraction (MWF) and g-ratio. We show that MWF should be computed using a three-compartment model in the corpus callosum, and the g-ratios obtained using three compartment models are consistent with previous reports. We provide results for other parameters, such as signal compartment frequency shifts.