Fig. 4: Histogram of the mean values for the diamagnetic and paramagnetic sources of all nuclei. Each Histogram shows values for the matrix nuclei in blue color and the core nuclei in red color. Note the lower values for the core nuclei in contrast to matrix nuclei indicating the more substantial contribution of diamagnetic and paramagnetic sources, i.e., iron, myelin, and calcium in the matrix nuclei.

Fig. 1: Thalamus QSM at 9.4 T. A) QSM map (scaled between: -215.0546 to 208.2119) B1) Negative values: diamagnetic sources (i.e., Myelin, Calcium, etc.) B2) Positive values (Paramagnetic sources, i.e., Iron, etc.) Both the values are in parts per billion, i.e., ppb (magnetic strength irrelevant). The depicted color code for all views is hot. The negative Myelin map shows a homogeneous distribution of negative QSM values across the thalamus; in contrast, the positive map shows increased QSM values mainly at the posterior, lateral, and intralaminar, midline nuclei.

Figure 1: Example images from one subject in the three orthogonal views. The red arrows indicate the genu of the corpus callosum where in some methods a common hyperintensity was observed. R1 – longitudinal relaxation rate, MTR – Magnetization Transfer Ratio, MPF – Macromolecular Proton Fraction, ihMTSat – Inhomogeneous magnetization transfer saturation, MWF – Myelin Water Fraction, IR-UTE – Inversion Recovery Ultra-Short Time-to-Echo.

Figure 3: Agreement and Bland-Altman plots for each method that show the reproducibility of each method. Colours indicate different tissue types: blue - brain structures, red - ventricles, yellow - gray matter, purple - white matter.

Across subjects average (A) perfusion-weighted (left) and perfusion (right) flatmaps. Subfield-specific averages, per subject and hemisphere, are shown in B for both perfusion-weighted (left) and perfusion metrics (right). Semi-transparent points show vertex-wise averages. Across subjects average perfusion-weighted signal (left) and standard variation (right), as a function of number of included runs and separated per subfield (colored lines) are displayed in C.

Correlation analyses between vertex-wise perfusion-weighted average across-subjects (A, y-axis), standard variation (B, y-axis) and distance (x-axes). Marginal plots show distribution for the metric displayed along that axis. Points are color-coded based on the diameter of most nearby vessel. Overlaid solid and dashed lines indicate corresponding best fits for each category of diameter.

Figure 2. Top row shows a zoomed-in image around Heschl's gyrus. Bottom row shows a zoomed-in image around the calcarine sulcus. Left column shows MP2RAGE UNI contrasts and right column shows T₂^* measurements obtained from ME-GRE.

Figure 3. A) Showing steps for generating our regions of interest centered around the primary auditory and visual cortices. B) showing steps for the depth metrics and B₀ alignment.

Fig. 1. The constructed fetal cortical surface atlases (left hemisphere) at different gestational weeks (GW). The number in the bracket indicates the subject number for construction the atlas at the time point. (a) The sulcal depth on the spherical surface. (b) The average convexity on the spherical surface. (c) The mean curvature on the spherical surface. (d) The sulcal depth on the inner cortical surface. (e) The average convexity on the inner cortical surface. (f) The mean curvature on the inner cortical surface. (g) The Desikan parcellation on the inner cortical surface.

Fig. 2. The cortical surface area of the atlases at different gestational weeks, in both the left and right hemisphere.

Figure 2. Tissue-based distribution of lag parameters for each respiratory design

(A) Refence anatomical axial slices in MNI space. (B-D) Averaged lag maps (seconds). The cumulative percent frequency (normalized to 100%) for the distribution of lag (seconds) is shown for each tissue which was extracted using the grey- (black) and white- (grey) matter mask displayed in (B), bottom right corner. A dotted red line was added to each histogram (B-D) at 40 seconds, for reference and comparison.

Figure 1. Grey- and white-matter average CVR lag times for each stimulus

Fig. 1. Flowchart of multi-atlas label fusion framework. Firstly, a non-linear registration is performed to align 3T individual to 7T QSM image space; then inverted deformation fields are applied onto 7T thalamic sub-nucleus parcellation map to generate individual estimation maps; finally, joint label fusion method fuses these parcellation map into 3T atlas space.

Fig. 3. Thalamic parcellation map on a representative slice of 3T atlas image shown in axial (top row), sagittal (middle row), and coronal (bottom row) views. The Schaltenbrand and Wahren histology atlas references are shown on the right.

Figure 4. Whole-brain 3D maps of Y_v, DBV, R′_2,nb, and χ_nb in the three orthogonal planes, obtained using the proposed qBOLD method. Note the expected contrast and physiologically plausible range in all four parameters, i.e., a clear depiction of gray/white matter differentiation, relatively homogeneous Y_v, and elevated R′_2,nb and χ_nb in the deep brain structures relative to cortical regions.

Figure 1. Schematic of the proposed, preliminary estimates-based qBOLD parameter mapping procedure.

Figure 2. Axial PD, T1, T2, T1ρ, T2*, QSM, and ∆B0 maps of three slices using the proposed method.

Figure 4. Synthetic PDw, T1w, T2w, T2*w, FLAIR, SWI, tSWI of three axial slices corresponding to Figure 2, as well as mIP and tmIP calculated from SWI and tSWI respectively using 16 axial slices. Contrast-weighted images are synthesized with good image quality, image contrasts, and venous structures.

Fig. 1. Conceptual schematic of magnetic resonance recording of local neuronal firings (mrLNF). A) Localization of local neuronal firing volumes defined by multiple individual coil sensitivities (shading areas) or by selective single-voxel excitation (dashed line). B) Intrinsic micro quantum sensors of nuclear spins (short arrows) immersed in ionic flow (neuronal current) inside neurons. C) Modulation of neuronal firing magnetic field B_n onto MR main field B₀. D) Recording of neuronal firing via MRI scanner.

Fig. 3. The mrLNF recordings on a 35-year-old female subject’s head at resting state at 3T. A) Localization of the mrLNF recordings on sodium MRI images. B) Raw FID signal at Channel 0. C) The mrLNF recordings from single FID acquisition, showing a fast action potential firing at Channel 0 (arrow, 2ms in duration, 648.6 μT in intensity). D) The mrLNF recordings from three consecutive FID acquisitions.

Visualization of hand-crafted nuclei depictions. (a) Overall visualization of the parcellation map in axial, sagittal and coronal views along with a 3D rendered view (label information only given on right hemisphere); (b) Labeled CN, NAC and VP on the sections of three views; (c) Labeled Pu, GPe and GPi on the sections of three views; (d) Labeled STN, RN, SNr and SNc on the sections of three views.

Pipeline of MNI space QSM atlas generation. (a) Special normalization for individual T1w and QSM image; and hybrid image generation. (b) Atlas space construction. (c) Guiding by T1w template, QSM atlas is projected into MNI space.

Figure 2: Exemplar skull-stripped T₁w template images (top row) and corresponding single subject coronal slices from images registered using multi-contrast registration (second row), single-contrast ANTs (third row), FLIRT (fourth row) and FNIRT (last row). Single- and Multi-contrast images match template neuro-anatomy more closely in skull-stripped data compared to other methods.

Figure 1: Exemplar T₁w template coronal images in a single subject (top row) to visually compare multi-contrast registration (second row), single-contrast ANTs (third row) and FLIRT (bottom row) in whole-head data (reference space). Overlay: T₁w template brain mask. Multi-contrast technique shows improved overlap between template and registered images. White arrows show underestimation of brain surface in single-contrast ANTs.

Fig. 1 Conventional GRE and 2-time accelerated SE-GRE image comparison under the same resolution with the zoomed-in comparison of the center region. The two images have similar SNR, and their SSIM value is above 0.99, indicating an extremely high resemblance for image quality in terms of signal strength, contrast, and structural similarity.

Fig.3 Standard GRE (Res. = 1 mm) and SE-GRE (Res. = 0.5 mm) images acquired within the same scan time and the zoomed-in comparisons of the bottom-left region and the top-right region. The comparison shows significant improvement for clarity of structural details, visibility of vessels, and edge sharpness between different tissues.

FIG 3. Representative 3D T1W multiplanar images with volumetric segmentation on a 3T scanner. [Left to right]: Sagittal, coronal, axial T1W images with SOC (scan time 5:01 min) on the top row and FAST-DL (scan time 2:37 min) on bottom row.

FIG 4. Representative axial 3D T1W images on a 3T scanner. [Left to right]: SOC (scan time 9:13 min), FAST (scan time 4:36 min), FAST-DL (scan time 4:36 min).

Figure 1. Scheme of the MIXTURE (Multi-Interleaved X-prepared tse with inTUitive RElaxometry)

T2-mapping was performed using T2-prepared 3D segmented turbo spin-echo (TSE) with variable refocusing pulse trains. Two images with different TE (TE = 0 and 50ms) were acquired with interleaved acquisition. To obtain FLAIR contrast, inversion time (TI) is adjusted to suppress cerebrospinal fluid.

Figure 3. Trigeminal nerve presentation of multi-echo TSE, MIXURE T2, and MIXTURE FLAIR

Arrows indicates bilateral trigeminal nerves. The prepontine cistern around the trigeminal nerve looks noisy on T2 maps of multi-echo TSE and MIXURE T2, whereas the noise is reduced on the T2 map of MIXURE FLAIR.

Figure 3: Four representative segmentation cases by VISIBLE-DM:

Segmented regions are shown in pink. Ground truth created by radiologists are also shown for comparison.

Figure 2: Model development: a) Annotation examples (white arrows), b) Models are developed with three kinds of input data, black and bright, black only and bright only images.

Figure 3: Comparison of SPs at different AFs. (a) shows FS k-space and SENSE reconstruction. (b) shows Poisson-disc SP at different AFs (4, 10, 20, 30) and the corresponding Low-rank reconstruction. (c) shows the optimized SP at the same AFs as (b) and resulting low-rank reconstructions. The improvement in performance is highlighted by the arrows.

Figure 5: Comparison shows improved performance for optimized SP. (a) shows NRMSE errors across iterations for k-space training and validation sets (b) shows NRMSE errors across iterations for image training and validation sets (c) shows lower k-space NRMSE errors at different AFs for the optimized SP compared to PD SP, both SP using low-rank reconstruction (d) shows lower image NRMSE errors at different AFs for optimized SP vs PD (e) the improvement of the optimized SP vs PD in T_1ρ mapping compared to FS reference at different AFs.

Figure 4: LV1 cognition-cortical volume saliences. A) Cortical volume saliences in TRS (red) and controls (green). Black lines: confidence intervals (CIs). Unreliable regions (CIs crossing zero) are grayed out. B) Cognitive saliences. y-axis: salience strength, x-axis: cognitive variables: PAL total errors (PAL-TE), stages completed (PAL-SC), IED interdimensional (IED-IS) and extradimensional shift (IED-ES), spatial-span length (SSP-SL), spatial working-memory strategy (SWM-S), total errors (SWM-TE). C) Reliable volume saliences projected onto a glass-brain.

Figure 2: Decomposition of multi-block cross-correlation matrix into additive components using MB-PLS-C framework (x-axis: cognitive measures, y-axis: regional cortical volumes). The first two components corresponding to the significant latent variables, LV1 and LV2, describe correlations in the latent space between structure and cognitive measures across patients and healthy controls. Four disjoint data blocks with multivariate measures of brain structure, cognition and covariates.

Figure 1: Mean control perfusion maps measured by ASL and 15O-water (whole brain CBF = 61.6 ± 12.4 ml/100g/min).

Figure 2: T-maps generated using relative and absolute CBF measured by ASL and ¹⁵O-water in 4 exemplary FTD patients. Areas in red (ASL) and blue (¹⁵O-water) indicate regions of significant hypoperfusion in the patient participant compared to the control group. sv = semantic variant, bv=behavioural variant, nfPPA = non-fluent primary progressive aphasia, psp = progressive supranuclear palsy.

Example data files for one subject, in T1-weighted space. Top panel shows the Harvard-Oxford Cortical atlas: 48 regions split into left and right hemisphere parcels. The middle panel shows CVR maps, not optimized for lag (No-Opt) and optimized for lag (Lag-Opt), for the BH+REST data segment. The bottom panel shows baseline CBF maps. Average CBF and CVR were obtained from each parcel.

The association between bCBF and CVR values in 96 cortical parcels, compared across each of the five data segments (columns), for CVR values with no lag optimization (No-Opt) and CVR values with lag optimization (Opt). The thick lines that connect filled dots represent the group mean, and the thin lines represent each individual subject. The shaded grey box represents which correlations were not significant at the single subject level, based on a critical value of 1.96 (p<0.05, two-tailed).

Figure 1: A) SG signal derived from the tracking of the ON position and the ON diameter for both eyes. The bins for two ON positions are marked in yellow with the reconstructed images being presented in Fig. 3. B) Scout image in axial orientation with visible ON (solid arrow), EOMS (dashed arrows) and orbital fat (dotted arrows) and two exemplary lines (green) on which the location of the ON is calculated and tracked.

Figure 3: Images resulting from reconstruction of profiles (A) without SG, (B) with SG from the bin for centre location of the ON and (C) with SG from the bin for ‘looking left’ as determined by the SG signal in Fig 1. While motion blur of the ON (solid arrow) is clearly visible in A, the image-based SG approach enables the reconstruction of images in various motion phases with reduced motion blur (solid arrows in B,C) and allows e.g. the visualization of an EOM contraction (striped arrow in C).

Comparison between the OEF obtained by QQ-CAT and QQ-CCTV in 6 stroke patients imaged between 6hrs and 10 days post stroke onset. QQ-CCTV generally shows more uniform OEF maps compared to QQ-CAT. In 7 and 9 days post-onset patients, low OEF areas in QQ-CCTV agree better with DWI-defined lesions.

Figure 1. Comparison between the OEF obtained by QQ-CAT and QQ-CCTV in a simulated stroke dataset. The numbers indicate RMSE (yellow) and MSD (white). On average, QQ-CCTV provide greater accuracy (MAE: 4.5 and 3.5%) and precision (MSD: 5.1 vs 1.7%). The OEF^_avg and OEF_std indicates the average and standard deviation OEF map among 5 trials, respectively.

Figure 1: R₂* behavior in relation to increasing angle of white matter fibers. Post mortem data recorded with b = 2000 s/mm² (red), b = 6000 s/mm² (green) and b = 9000 s/mm² (yellow) are marked with dots and in vivo data recorded with b = 800 s/mm² (blue) and b = 2000 s/mm² (red) are marked with triangles. The shaded areas represent the 95% confidence interval of the respective data.

Figure 2: Boxplots for MD (left) and FA (right) of different b-values used for in vivo (red) and post mortem (blue) scans. Statistical significant differences < 0.05 are marked with *, < 0.01 with + and < 0.001 with x.

Figure 4. GLM for a 4 block-design scan. Voxels significantly modulated by stimulus presentation in the occipital cortex with (p < 0.001) after cluster-based correction for multiple comparisons. Statistical analysis was performed with a GLM from 4 block-design scans (11 12-sec blocks of visual stim interleaved with 10 12-sec blocks of rest). The GLM regressors were formed by convolution of boxcar representing stimulus presentation with a simple gamma function $$$g(t) = t^q \cdot e^{-t} / (q^q \cdot e^{-q}), q=4 $$$

Figure 3. Channel dependent impact of NORDIC applied before GRAPPA reconstruction with R=3 for a single channel covering a sagittal slice through the occipital lobe.

Table 2. Effect of adjusting for hardware changes. BPF is unitless, TD units are 10-3 mm2/sec. Values are change per year, with 95% CI’s in parentheses. Models are: no adjustment for hardware changes (Original), data acquired after a hardware change excluded (Exclude), hardware change treated as a binary yes/no time-dependent covariate (Binary) and type of hardware change is a time-dependent covariate (Type). Lower AIC indicates a better fit, with a change of 2 being substantial.

Table 4. Impact of scanner change on overall outcomes. Values are differences in treatment effect for BPF or TD by a particular type of hardware change versus no change. The GE HDxt to GE MR750 change could not be analyzed because only one subject was affected.

Fig. 3: (A) - (C): PSR maps of a healthy volunteer (male, 36 yrs) under the CS-SENSE factors of 3, 8, and 12, respectively; (D) - (F) show the corresponding R_1f maps.

Fig. 1: (A) - (C) PSR maps of the BSA phantoms under the CS-SENSE factor of 3, 8, and 12 respectively, the phantoms' concentrations are labeled on (A). The corresponding R_1f maps under each CE-SENSE factor are shown in (D) - (F).

Figure 3: Measured versus reference T1 (left) and T2 (right). Colour of the symbols indicates vendor, the shape indicates acceleration used, and filled vs hollow indicates 2D vs 3D respectively.

Figure 4: Difference between measured and reference T1 (left) and T2 (right) values. Colour of the symbols indicates vendor, the shape indicates acceleration used, and filled vs hollow indicates 2D vs 3D respectively.

Figure 1. Differential identifiability and distance metrics for original T1 data and for T1 data with added noise. (A) displays computed I_diff values for original T1 images (black) and T1 images with added noise (red), reconstructed with the full range of possible principle components (PCs). (B) displays D_self (blue) and D_others (green) values for original T1 images and T1 images with added noise, reconstructed with varying number of PCs.

Figure 3. T1 anatomical images associated with original and added noise datasets at full (36 PC) reconstruction and optimized I_diff (18 PC) reconstruction.

Figure 1. Image-to-image translation using FPG for two MS subjects for three slice positions (in rows). The input images (MS) are translated to the healthy domain (HC’). The absolute difference between the input and the translated images reveals the disease-related voxels, which are overlaid on the original input images. Green indicates an increase and pink a decrease in intensity.

Figure 2. Image-to-image translation using FPG on two MS subjects for one slice position. The input images (MS) are translated to the healthy domain (HC’) (upper row) and to the diseased domain (MS’, lower row). The absolute difference between the input and the translated images is threshold-limited to highlight disease-related voxels, which are overlaid on the input images. Green indicates an increase and pink a decrease in intensity.

Figure 1. (top panel) the signal model for the QALAS sequence is shown with M₀=1, describing the effect of different flip angles (different colours), and the corresponding measured data. (middle panel) QALAS data from an MS patient showing the signal maps for the individual dynamics. (bottom row) Corresponding QALAS data of a relaxation phantom.

Figure 3. Bland-Altman plots of the differences between R1 and R2-measurements using the implemented QALAS sequence, compared to gold standard methods. QALAS measurements with FLASH flip angle of 4° (yellow) and green for FLASH flip angle of 10° (green).

Neuroimaging characteristics. Mean volume (in mm3), standard deviation and percent difference.

Example of hippocampal subfields segmentation in a subject with SCD. a: coronal slice of the T2-weighted image, acquired at 7T with resolution 0.375x0.375x1.5mm2; b,c: zoom-in images showing details of the hippocampus structure, subject right and left, respectively; d,e: hippocampus subfield segmentations overlaying the T2-weighted image, subject right and left, respectively; f: 3D reconstruction of the hippocampal subfield segmentations.

Figure 3. Results of the k-means clustering analysis per state. A. Cluster centroids for each state. B. The strongest 5% of the functional connectivity matrix in each state. Red lines represent positive functional connectivity, and blue lines represent negative functional connectivity. VN = visual network; SMN = sensorimotor network; DMN = default mode network; AN = auditory network; LFPN = left frontoparietal network; RFPN = right frontoparietal network; CER = cerebellum; DAN = dorsal attention network; ECN = executive control network.

Figure 5. Between-group comparison in temporal properties of functional connectivity states and the partial correlation analysis results. Dot plots show individual data points (squares), averages (transverse lines), and standard deviations (vertical lines) of the mean fractional windows (A), mean dwell time (B), and total number of transitions (C). The transverse lines represent averages, and the vertical lines represent standard deviations. D-E. Partial correlation analysis results. *p < 0.05. TMT-A = Trail Making Test A; SDMT = Symbol Digit Modalities Test

Figure 1. Pearson’s correlation coefficients of ASL CBF ratio and uncorrected DSC rCBV in enhancing glioma.

Figure 2. Pearson’s correlation coefficients of ASL CBF ratio and uncorrected DSC rCBV in non-enhancing glioma.

Figure 1: Brain regions with reduced cerebral blood flow in T2DM patients over control subjects. The sites with reduced cerebral blood flow in T2DM patients were included the bilateral prefrontal cortices (a, b), right insula (c), bilateral cingulate (d, h), bilateral cerebellum (e, g), and right thalamus (f). All images are in neurological convention (L = left; R = right). Color bar indicates t-statistic values.

Figure 3: Cognition showed positive associations with cerebral blood flow in T2DM patients. Positive correlations appeared between MoCA scores and cerebral blood flow of the bilateral prefrontal cortices (a, c), right thalamus (b), right putamen (d), bilateral hippocampus (e, i), bilateral cerebellum (f, g), right amygdala (h), and bilateral basal forebrain (j, k). Figure conventions are same as in Figure 1.

Fig. 1 Schematic diagram of the ROI measurement in various regions of the brain

Fig. 2 Bar chart comparison of the iron content measurement results of various brain regions in each age group

Axial (A) and corononal (B) heat map of segments differences between controls and CDH subjects. Percent change for CDH subjects when compared to controls is displayed. Segments in blue represent volume gain, while segments in red correspond to volume loss. White segments were not significantly different from controls.

(A)T2-weighted super-resolution volume reconstruction with inter-slice motion correction, and intensity normalization of a T2-weighted fetal acquisition (29 week-old fetus). (B) Propagation of fetal atlas labels. (C) 3D volumetric rendering of propagated labels.

Figure 4. Intraclass coefficient correlations between two examiners for the sciatic nerve size and signal measurements. Bold represents ICC> 0.601. T2-FFE, DN, and DTI show substantial agreements in the long diameter, while T2WI, 3D-iNerveVIEW, and DTI show substantial agreements in the short diameter. Regarding the signal measurements, 3D-iNerveVIEW shows substantial to almost perfect agreements in signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and contrast ratio (CR).

Figure 5. Problems with 3D-iNerveVIEW. Case1: By suppressing the signal of the vessel, the boundary between nerves and blood vessels becomes unclear. In addition, the vessel walls running in parallel in the long diameter direction of the sciatic nerve become high signals, which reduces visibility. Case2: The vessels are misidentified and measured as nerves when the signal of the vessels running parallel to the long diameter of the sciatic nerve cannot be suppressed.

Figure 2: R₂^* as a function of the WM fiber angle averaged over the in vivo subjects (red) and the post mortem subjects (blue) fitted with the absolute values (A) and normalized to the global mean WM R₂^* (B). The shaded areas represent the 95% CI of the measured data.

Figure 1: Representative maps of R₂^* (A) and fiber angle θ (B) of two post mortem subjects with brain temperatures of 5.6°C (top row) and 14.9 °C (middle row) and one in vivo subject shown in the bottom row.

Figure 1: The proposed image processing pipeline. UTE scans at three different flip angles (18, 12, and 6 degrees) are fit with the following signal model. The resulting parameter map corresponding to the ultrashort-T₂* fractional component is then used to isolate values in various brain ROIs.

Figure 3: Split violin plots comparing the distributions of the ultrashort-T₂* component fraction in various brain ROIs across 4 healthy volunteers. The dashed lines represent the interquartile range of each distribution.

Figure 1. Plots of significant findings between DTI metrics and demographic information. Top left: Total FA injury burden versus Age; Top Right: Total FA injury burden versus PCSS Score; Bottom Left: Total RD injury burden versus Age; Bottom Right: A paired matrix plot indicating the distribution of each metric and the correlation between metrics.

Table 1. Results for the multilinear regression of demographic and DTI metrics.

Figure 2: Combined surviving components (components 1, 4 and 42, 98% of r² retained) overlaid on the MNI152, 2mm brain template. Negative associations with MoCA are depicted in blue, and positive in red, with the ECN map¹¹ in yellow.

Figure 4: Colour-coded graph depicting the interrelatedness between any pairs of nodes. Positive associations are depicted with red edges, and negative with blue. The circle represents the mid-axial slice of the brain.

Figure 2: Network based statistics (NBS) under the contrast [-1, 1] for OSA and healthy control group connectivity comparison.

Figure 3: Scatter plot of significant effects found in data collected from behavioral versus brain measure comparison for both healthy and OSA subjects under threshold value of 0.35.

Figure 4a: The venous blood oxygen saturation (SvO2) alterations across gestational age (GA) of the normal group: SvO2 = -0.004223*GA + 0.8783. Figure 4b: SvO2 alterations across gestational age between the congenital heart disease (CHD) group (SvO2 = -0.00514*GA + 0.942) and GA-matched normal group (SvO2 = -0.01205*GA + 1.087) and their corresponding fitting curve. Red circles and red curves represent the CHD group; blue circles and blue curves represent the normal group.

Figure 5: Venous blood oxygen saturation (SvO2) values in the congenital heart disease (CHD) group and gestational age (GA)-matched normal group, along with their corresponding mean and 95% CI SvO2 values. In the analysis of covariance, the SvO2 value showed a significant difference between the CHD (80.8%±4.6%) group and the GA-matched normal group (75.7%±8.0%) after the effects of GA were excluded (p=0.038).

Figure 3: Spatial correlation across 28 gray matter VOIs. Each cross indicates the mean parameter value in a specific Brodmann area averaged across subjects. Strongest spatial correlation was found between PET- and mqBOLD-OEF (A) with r=0.69 (p<0.05). Correlation between PET-OEF and DSC-OEC (B) was also strong, but slightly weaker with r=0.50 (p<0.05). Similarly, good correlation was also found between both MRI techniques (C) with r=0.57 (p<0.05). Note that MRI and PET data were acquired in similar cohorts, but different subjects.

Figure 2: Exemplary PET-OEF, MRI-OEF and MRI-OEC data in four different slices. PET-OEF (left column) is quite homogenous across the entire brain, without contrast between gray (GM) vs. white matter (WM). MRI-based OEF (central column) and OEC (right column) show some similarity with PET. However, both parameter maps show an artifactual GM/WM contrast with elevated WM values. While OEF values in GM seem similar between PET and mqBOLD, OEC values based on the CFIN-model are generally lower.

The custom designed automatic delivery system of CO₂-CVR included mass flow controllers with medical graded gas mixtures and air, LabView for device control, CO₂ analyzer with PoweLab and Chart 5 with synchronization with radiofrequency pulse of MRI.

Cortical specificity of CO₂-CVR showed highest t values of one-sample t test using carbon response function (see text for details) as 14.53 at ventral portion of Brodmann area 23 (blue). And cortices with CO₂-CVR larger than 13 were labelled as green and red for 13 < t values < 14 and t values >14, respectively.

Figure 1: Automatic processing pipeline: The ICBM 2009a nonlinear symmetric MNI template was non-linearly registered to the subject’s T1w space (using Vida_64Ch as reference). Inter/Intra-run co-registrations were performed (T1wRef-to-T1w, as well as T1w-to-R2*). Tissue segmentation was achieved using SPM12 software using the default brain probability maps. Labels were propagated to R2*/QSM space and restricted to WM/GM tissue types. WM/GM lobes and cerebellum were extracted, as well as four deep grey nuclei and the brainstem prior ROI analysis on the quantitative maps.

Reproducibility across MR system types (Verio and Vida), receive head coils (20, 32 and 64 channels) and sequence parameters. Note that the measured variability on the Vida with different head coils was not different than the intra-run metric variations on the Verio system. The signal drop caused by the increased acceleration factor from 3 to 4 was not reflected in the R2* and QSM estimations, with values in the same range as the other protocols. The central table reports mean and standard deviation per segmented region across all measurements.

AD

amyMCI

Fig.1 Significant edge list for positive associations with d’-total scores. The nodes with its total number of connections are also indicated.

Fig.2 Significant edge list for positive associations with d’s computed from sequential processing component (comprising Melody, Standard Rhythm, Embedded Rhythm and Accent). The nodes with its total number of connections are also indicated.

Fig. 1 Top row: magnitude of corrected, uncorrected and lower-resolution corrected (isotropic 0.6 mm) T₂*-weighted MRI from one subject. The yellow arrow points to the hypointensive Line of Gennari in the primary visual cortex. Bottom row: the corresponding susceptibility-induced off-resonance frequency maps at 7 T. The Line of Gennari appears as a positive-frequency band (note darker equals more positive).

Fig. 2 Normalized root mean square error (NRMSE) between two repeated T₂*-weighted scans for each subject. Red line represents the expected NRMSE caused by thermal noise.

T1Mapping maps in the fetal brain (A–D). Examples of ROIs for DTI are shown in white color (1–2). The area of the ROI was adjusted appropriately according to the gestational week and anatomical structures. Regions of interest: 1, Thalamus; 2, Corticospinal fibers. A and B are T1Mapping images of the same fetus; And gestation ages was 28 week. C and D are T1Mapping images of the same fetus; And gestation ages was 36 week.

T1 relaxation time trajectories in Thalamus for 17 subjects. T1 had good correlation with gestational age. A regression line is shown along with 95% confidence intervals (dotted lines).