Search

The vast majority of studies using functional magnetic resonance imaging (fMRI) are analysed on the group level. Standard group-level analyses, however, come with severe drawbacks: First, they assume functional homogeneity within the group, building on the idea that we use our brains in similar ways. Second, group-level analyses require spatial warping and substantial smoothing to accommodate for anatomical variability across subjects. Such procedures massively distort the underlying fMRI data, which hampers the spatial specificity. Taken together, group statistics capture the effective overlap, rendering the modelling of individual deviations impossible-- a major source of false positivity and negativity. The alternative analysis approach is to leave the data in the native subject space, but this makes comparison across individuals difficult. Here, we propose a new framework for visualizing group-level information, better preserving the information of individual subjects. Our proposal is to limit the use of invasive data procedures such as spatial smoothing and warping and rather extract regional information from the individuals. This information is then visualized for all subjects and brain areas at one glance – hence we term the method brainglance. Additionally, our method incorporates a means for clustering individuals to further identify common traits. We showcase our method on two publicly available data sets and discuss our findings.

Long-TE gradient recalled-echo (GRE) scans are prone to phase artifacts due to B0 inhomogeneity. We propose a learning-based approach that does not rely on navigator readouts and allows to infer phase error offsets directly from corrupted data. Our method does not need to be pre-trained on a database of medical images that match a contrast/acquisition protocol of the input image. A sufficient input is a raw multi-coil spectrum of the image that needs to be corrected. We train a convolutional neural network to predict phase offsets for each k-space line of a 2D image. We synthesize training examples online by reconvolving the corrupted spectrum with point spread functions (PSFs) of the coil sensitivity profiles and superimposing artificial phase errors, which we attempt to predict. We evaluate our approach on “in vivo” data acquired with GRE sequence, and demonstrate an improvement in image quality after phase error correction.

Purpose: Knowledge of the positions of field probes in an NMR field camera is necessary for monitoring the B0 field. The typical method of estimating these positions is by switching the gradients with known strengths and calculating the positions using the phases of the FIDs. We investigated improving the accuracy of estimating the probe positions and analyzed the effect of inaccurate estimations on field monitoring. Methods: The field probe positions were estimated by 1) assuming ideal gradient fields, 2) using measured gradient fields (including nonlinearities), and 3) using measured gradient fields with relative position constraints. The fields measured with the NMR field camera were compared to fields acquired using a dual-echo gradient recalled echo B0 mapping sequence. Comparisons were done for shim fields from second- to fourth-order shim terms. Results: The position estimation was the most accurate when relative position constraints were used in conjunction with measured (nonlinear) gradient fields. The effect of more accurate position estimates was seen when compared to fields measured using a B0 mapping sequence (up to 10–15 more accurate for some shim fields). The models acquired from the field camera are sensitive to noise due to the low number of spatial sample points. Conclusion Position estimation of field probes in an NMR camera can be improved using relative position constraints and nonlinear gradient fields.

Functional localizers are particularly prevalent in functional magnetic resonance imaging (fMRI) studies concerning face processing. In this study, we extend the knowledge on face localizers regarding four important aspects: First, activation differences in occipital and fusiform face areas (OFA/FFA) and amygdala are characterized by increased activation while precuneus and medial prefrontal cortex show decreased deactivation to faces versus control stimuli. The face-selective posterior superior temporal sulcus is a hybrid area exhibiting increased activation within its inferior and decreased deactivation within its superior part. Second, the employed control stimuli can impact on whether a region is classified in group analyses as face-selective or not. We specifically investigated this for recently described cytoarchitectonic subregions of the fusiform cortex (FG-2/FG-4). Averaged activity across voxels in FG-4 was stronger for faces than objects, houses, or landscapes. In FG-2, averaged activity was only significantly stronger in comparison with landscapes, but small peaks within this area were detected for comparison versus objects and houses. Third, reproducibility of individual peak activations is excellent for right FFA and quite good for right OFA, whereas within all other areas it was too low to provide valid information on time-invariant individual peaks. Finally, the fine-grained spatial activation patterns in right OFA and FFA are both time-invariant within each individual and sufficiently different between individuals to enable identification of individual participants with near-perfect precision (fMRI fingerprinting).

The human brain is known to contain several functional networks that interact dynamically. Therefore, it is desirable to analyze the temporal features of these networks by dynamic functional connectivity (dFC). A sliding window approach was used in an event-related fMRI (visual stimulation using checkerboards) to assess the impact of repetition time (TR) and window size on the temporal features of BOLD dFC. In addition, we also examined the spatial distribution of dFC and tested the feasibility of this approach for the analysis of interictal epileptiforme discharges. 15 healthy controls (visual stimulation paradigm) and three patients with epilepsy (EEG-fMRI) were measured with EPI-fMRI. We calculated the functional connectivity degree (FCD) by determining the total number of connections of a given voxel above a predefined threshold based on Pearson correlation. FCD could capture hemodynamic changes relative to stimulus onset in controls. A significant effect of TR and window size was observed on FCD estimates. At a conventional TR of 2.6 s, FCD values were marginal compared to FCD values using sub-seconds TRs achievable with multiband (MB) fMRI. Concerning window sizes, a specific maximum of FCD values (inverted u-shape behavior) was found for each TR, indicating a limit to the possible gain in FCD for increasing window size. In patients, a dynamic FCD change was found relative to the onset of epileptiform EEG patterns, which was compatible with their clinical semiology. Our findings indicate that dynamic FCD transients are better detectable with sub-second TR than conventional TR. This approach was capable of capturing neuronal connectivity across various regions of the brain, indicating a potential to study the temporal characteristics of interictal epileptiform discharges and seizures in epilepsy patients or other brain diseases with brief events.

Purpose Phase artifacts due to B0 inhomogeneity can severely degrade the quality of MR images. The artifacts are particularly prominent in long-TE scans and usually appear as ghosting and blur. We propose a retrospective phase correction method based on autofocusing. The proposed method uses raw data acquired with standard imaging sequences, and does not rely on navigators or external measures of field inhomogeneity. Methods We formulate and solve the optimization problem, where we seek the latent phase offsets that are associated with an optimal value of the image quality measure that is evaluated in the spatial domain. As a quality measure we use entropy computed on spatial image gradients. We propose two types of objective function, both compatible with parallel imaging and accelerated image acquisition. Results We evaluate the method on both synthetic and real data. In real data case we evaluate the performance on a range of sequences and images acquired with different acceleration factors. The experimental results demonstrate that our method is capable of minimizing ghosting artifacts and that the quality of the output images is similar to navigator-based reconstructions. Conclusion The presented technique can be alternative to or complement navigator-based methods, and is able to improve images with severe phase artifacts from all standard imaging sequences.

Basierend auf Monte Carlo Simulationen wurden oxygenierungs- abhängige Signaländerungen bei balanced SSFP, Gradienten- und Spinechosequenzen untersucht. Signaländerungen wurden für künstliche Zylinder und reale neurovaskuläre Netzwerke aus dem Kortex der Maus mit 1μm Auflösung berechnet. Signaländerungen als Funktion des Gefäßdurchmessers, dem Blutvolumen, der Gefäßorientierung zum Magnetfeld sowie das Verhältnis der intra- und extravaskularen Beiträge wurden berechnet. Die Resultate zeigen, dass bSSFP eine hohe Sensitivität sowohl für intra- als auch extravaskuläre Komponenten zeigt. Außerdem zeigen Gradientenechos und bSSFP, aber weniger SE, eine starke Signaländerung mit der Orientierung der Gefäße zum äußeren Magnetfeld.

Purpose The purpose of this work was to develop a fast and efficient MRSI-FID acquisition scheme and test its performance in vivo. The aim was to find a trade-off between the minimal total acquisition time and signal-to-noise ratio of the acquired spectra. Methods Measurements were performed on a 9.4 Tesla system. Sequence optimization included redesign of water suppression, optimization of the sequence gradients, and improvement of the sampling efficiency by minimizing the read-out time. This resulted in an acquisition time of 2:47 and 22:13 minutes for 2D (TR = 57 ms; 3-mm in-plane resolution) and 3D MRSI (TR = 57 ms; 16 slices; 3-mm isotropic resolution), respectively. Results Despite strong T1 weighting and first-order phase problems, it was possible to obtain spectra of an acceptable quality. The average line width calculated for the tCr peak across the entire field of view was 26.9 ± 9.6 Hz for 2D and 30.0 ± 11.3 Hz for 3D MRSI. In 3D measurements, the percent fraction of voxels fitted with Cramer-Rao lower bounds below 10 was 53.3 ± 4.1, 63.4 ± 8.4, and 81.0 ± 2.9 for Glu, tCr, and tNAA, respectively. Conclusion Considering the typically long duration of high-resolution MRSI, the proposed technique may be of interest for clinical applications and/or studies that focus on following the biochemistry of dynamic processes.

PURPOSE Investigation of the feasibility to perform high resolution quantitative sodium imaging at 9.4 Tesla (T). METHODS A proton patch antenna was combined with a sodium birdcage coil to provide a proton signal without compromising the efficiency of the X nucleus coil. Sodium density weighted images with a nominal resolution of 1 × 1 × 5 mm(3) were acquired within 30 min with an ultrashort echo time sequence. The methods used for signal calibration as well as for B0 B1 and off resonance correction were verified on a phantom and five healthy volunteers. RESULTS An actual voxel volume of roughly 40 µL could be achieved at 9.4T while maintaining an acceptable signal to noise ratio (8 for brain tissue and 35 for cerebrospinal fluid). The measured mean sodium concentrations for gray and white matter were 36?±?2 and 31?±?1 mmol/L of wet tissue which are comparable to values previously reported in the literature. CONCLUSION The reduction of partial volume effects is essential for accurate measurement of the sodium concentration in the human brain. Ultrahigh field imaging is a viable tool to achieve this goal due to its increased sensitivity. Magn Reson Med 2014. © 2014 Wiley Periodicals Inc.

Purpose To assess the impact of colored noise on statistics in event-related functional MRI (fMRI) (visual stimulation using checkerboards) acquired by simultaneous multislice imaging enabling repetition times (TRs) between 2.64 to 0.26 s. Methods T-values within the visual cortex obtained with analysis tools that assume a first-order autoregressive plus white noise process (AR(1)+w) with a fixed AR coefficient versus higher-order AR models with spatially varying AR coefficients were compared. In addition, dependency of T-values on correction of physiological noise (respiration, heart rate) was evaluated. Results Optimal statistical power was obtained for a TR of 0.33 s, but T-values as obtained by AR(1)+w models were strongly dependent on the predefined AR coefficients in fMRI with short TRs which required higher-order AR models to achieve stable statistics. Direct estimation of AR coefficients revealed the highest values within the default mode network while physiological noise had little influence on statistics in cortical structures. Conclusion Colored noise in event-related fMRI obtained at short TRs originates mainly from neural sources and calls for more sophisticated correction of serial autocorrelations which cannot be achieved with standard methods relying on AR(1)+w models with globally fixed AR coefficients.

Purpose This work investigates the feasibility of high-resolution functional imaging of the human brain using passband balanced steady state free precession (SSFP) at 9.4 Tesla (T). To this end, the temporal signal stability, blood-oxygen-level-dependent (BOLD)-related signal changes and sensitivity to frequency offsets were evaluated. Methods Three-dimensional slab selective and nonselective balanced SSFP have been implemented with minimized repetition time and high temporal resolution using parallel imaging, partial Fourier acquisition and elliptical scanning. Using a volume repetition time of approximately 3 s, a visual checker board stimulation was applied for 6 min. Temporal signal stability of balanced SSFP and BOLD response-related signal changes and sensitivity to frequency changes were evaluated. Results Activation could be detected in all volunteers with BOLD-related signal changes from 3 to 6. At 1 mm isotropic resolution, the thermal noise SNR0 was 67 and the total temporal noise variation tSNR was 45 supporting a very high signal stability of balanced SSFP. No significant changes of activation at different offresonance frequencies were detected. Conclusion High spatial and temporal resolution balanced SSFP at 9.4T to detect functional activation is feasible. Activation patterns and signal changes are stable and reproducible across subjects within the visual cortex, and comparable to reported values of SE-EPI at 7T and 9.4T.

The formation of transient networks in response to external stimuli or as a reflection of internal cognitive processes is a hallmark of human brain function. However, its identification in fMRI data of the human brain is notoriously difficult. Here we propose a new method of fMRI data analysis that tackles this problem by considering large-scale, task-related synchronisation networks. Networks consist of nodes and edges connecting them, where nodes correspond to voxels in fMRI data, and the weight of an edge is determined via task-related changes in dynamic synchronisation between their respective times series. Based on these definitions, we developed a new data analysis algorithm that identifies edges that show differing levels of synchrony between two distinct task conditions and that occur in dense packs with similar characteristics. Hence, we call this approach ""Task-related Edge Density"" (TED). TED proved to be a very strong marker for dynamic network formation that easily lends itself to statistical analysis using large scale statistical inference. A major advantage of TED compared to other methods is that it does not depend on any specific hemodynamic response model, and it also does not require a presegmentation of the data for dimensionality reduction as it can handle large networks consisting of tens of thousands of voxels. We applied TED to fMRI data of a fingertapping and an emotion processing task provided by the Human Connectome Project. TED revealed network-based involvement of a large number of brain areas that evaded detection using traditional GLM-based analysis. We show that our proposed method provides an entirely new window into the immense complexity of human brain function.

Purpose Measurement of the cerebral metabolic rate of oxygen (CMRO2) by means of direct imaging of the 17O signal can be a valuable tool in neuroscientific research. However, knowledge of the longitudinal and transverse relaxation times of different brain tissue types is required, which is difficult to obtain because of the low sensitivity of natural abundance H217O measurements. Methods Using the improved sensitivity at a field strength of 16.4 Tesla, relaxation time measurements in the rat brain were performed in vivo and postmortem with relatively high spatial resolutions, using a chemical shift imaging sequence. Results In vivo relaxation times of rat brain were found to be T1 = 6.84 ± 0.67 ms and T2* = 1.77 ± 0.04 ms. Postmortem H217O relaxometry at enriched concentrations after inhalation of 17O2 showed similar T2* values for gray matter (1.87 ± 0.04 ms) and white matter, significantly longer than muscle (1.27 ± 0.05 ms) and shorter than cerebrospinal fluid (2.30 ± 0.16 ms). Conclusion Relaxation times of brain H217O were measured for the first time in vivo in different types of tissues with high spatial resolution. Because the relaxation times of H217O are expected to be independent of field strength, our results should help in optimizing the acquisition parameters for experiments also at other MRI field strengths.

PURPOSE: Measurement of the cerebral metabolic rate of oxygen (CMRO2 ) by means of direct imaging of the 17 O signal can be a valuable tool in neuroscientific research. However, knowledge of the longitudinal and transverse relaxation times of different brain tissue types is required, which is difficult to obtain because of the low sensitivity of natural abundance H2 17 O measurements. METHODS: Using the improved sensitivity at a field strength of 16.4 Tesla, relaxation time measurements in the rat brain were performed in vivo and postmortem with relatively high spatial resolutions, using a chemical shift imaging sequence. RESULTS: In vivo relaxation times of rat brain were found to be T1 = 6.84 +/- 0.67 ms and T2 * = 1.77 +/- 0.04 ms. Postmortem H2 17 O relaxometry at enriched concentrations after inhalation of 17 O2 showed similar T2 * values for gray matter (1.87 +/- 0.04 ms) and white matter, significantly longer than muscle (1.27 +/- 0.05 ms) and shorter than cerebrospinal fluid (2.30 +/- 0.16 ms). CONCLUSION: Relaxation times of brain H2 17 O were measured for the first time in vivo in different types of tissues with high spatial resolution. Because the relaxation times of H2 17 O are expected to be independent of field strength, our results should help in optimizing the acquisition parameters for experiments also at other MRI field strengths. Magn Reson Med, 2015. (c) 2015 Wiley Periodicals, Inc.

Obesity-related structural brain alterations point to a consistent reduction in gray matter with increasing body mass index (BMI) but changes in white matter have proven to be more complex and less conclusive. Hence, more recently diffusion tensor imaging (DTI) has been employed to investigate microstructural changes in white matter structure. Altogether, these studies have mostly shown a loss of white matter integrity with obesity-related factors in several brain regions. However, the variety of these obesity-related factors, including inflammation and dyslipidemia, resulted in competing influences on the DTI indices. To increase the specificity of DTI results, we explored specific brain tissue properties by combining DTI with quantitative multi-parameter mapping in lean, overweight and obese young adults. By means of multi-parameter mapping, white matter structures showed differences in MRI parameters consistent with reduced myelin, increased water and altered iron content with increasing BMI in the superior longitudinal fasciculus, anterior thalamic radiation, internal capsule and corpus callosum. BMI-related changes in DTI parameters revealed mainly alterations in mean and axial diffusivity with increasing BMI in the corticospinal tract, anterior thalamic radiation and superior longitudinal fasciculus. These alterations, including mainly fiber tracts linking limbic structures with prefrontal regions, could potentially promote accelerated aging in obese individuals leading to an increased risk for cognitive decline., Highlights • We combined DTI with quantitative multi-parametric mapping (qMPM). • Obesity was associated with a reduction in myelin displayed by decreased R1. • Obesity was associated with increased water content (decreased R1 and increased PD*). • DTI revealed mainly alterations in mean and axial diffusivity with obesity.

In general MR image contrast is expected to be resolution independent but a pronounced loss of contrast is observed between fluids and tissues with contemporary musculoskeletal protocols (typical inplane resolution << 1 mm) using nonbalanced steady state free precession such as double echo steady state. For nonbalanced steady state free precession diffusion sensitivity increases with increasing spoiler moments which increase with decreasing voxel size suggesting diffusion damping as the major cause for the observed contrast variation. This is confirmed by simulations and measurements indicating that for fluids diffusion effects become apparent already for resolutions ?x < 1 mm whereas tissues typically require ?x < 200 µm. Gradient spoiling however is generically not minimized but frequently applied along the readout direction. For anisotropic steady state free precession scans the loss of contrast between fluids and tissues from diffusion can thus be minimized by simply moving the spoiler gradients to the lowest resolution direction. Copyright © 2012 Wiley Periodicals Inc.

Purpose Standard gradient-echo sequences are often prohibitively slow for inline image-weighted imaging as long echo times prolong the repetition time of the sequence. Echo-shifting offers a way out of this dilemma by allowing an echo time that exceeds the repetition time. The purpose of this work is to present a gradient-echo sequence that is optimized for multislice inline image-weighted imaging applications by combining echo-shifting with an interleaved slice excitation order. Theory and Methods This combined approach offers two major advantages: First, it combines the advantages of both concepts, that is, echo time and pulse repetition time can be significantly increased without affecting scan time. Second, there is no echo-shifting related signal loss associated with this concept as only a single radiofrequency pulse is applied per pulse repetition time and slice. Results A 9.4 Tesla high-resolution inline image-weighted anatomical brain scan of the proposed sequence is compared to a standard gradient-echo. Furthermore, results from 9.4 Tesla blood oxygen level dependent functional magnetic resonance imaging experiments with an in-plane resolution of 0.8 × 0.8 mm2 are presented. Conclusion The proposed sequence allows for efficient generation of inline image-weighted contrast by combining echo-shifting with an interleaved slice excitation order.

Sodium imaging is challenging due to its inherently low signal-to-noise ratio. However, the sensitivity can be increased considerably if data acquisition is performed at ultrahigh field with optimized MR coils and sequences. In this work, two novel coil setups for multi-nuclear sodium/proton imaging at 9.4 Tesla as well as some in vivo applications are presented.