Fig1.Neuroradiologist outlined 5 regions of interest (ROI)(A).Motion correction was done prior to B0 calculation. The signal % at 2 ppm was determined when B0 variation is lower than 1 ppm (B). The MTRasym at2 ppm represents the GlucoCEST signal (C). Dynamic DRY, INFUSION, CONTRAST acquired 20, 35 and 130 images respectively in 3 phases, each with 5 images, ranging from +3 to -1 ppm (D). Average of the 3 DRY phases generated a reference signal Sbase. INFUSION and CONTRAST acquired Sn images (E).

Fig2.Mean MTRasym values from 15 patients with brain tumours were significantly higher in ROI 4 than in ROI 5 (ANOVA and TukeyHSD test) (A). The mean MTRasym values from ROI 1 to3 were significantly different from ROI 5 in the same 15 patients (ANOVA and Dunnetttest) (B).

Figure 2. Quantitative maps estimated using MT modeling with standard and proposed R1b in a healthy volunteer. (a) MPF and R1f. As predicted by simulations (Fig. 1b), standard R1f is dominated by macromolecular content and therefore resembles MPF. The proposed R1f is much more uniform, likely due to removing the effects of MPF. Note slightly elevated values in the basal ganglia, likely due to iron accumulation. (b) Error between the proposed and standard MPF. Note its high variability with macromolecular content and B1 field (c), also consistent with simulations (Fig. 1a).

Figure 4. Quantitative mapping in an MS subject. (a) Proposed R1f is elevated in deep GM. The increase is consistent with the known distribution of iron in these regions, including areas with high (globus pallidus, #1), medium (putamen, #2), and low (thalamus, #3) iron content. Note elevated R1f in areas affected by diffuse lesional changes (#4) (as seen on T2-FLAIR and MPF). (b) The R1f is increased on the rim of the heavily demyelinated lesion (as revealed by MPF) (#1) and in WM areas affected by more diffuse disease (#2). (c) Example of a lesion without noticeable increase in R1f.

Figure 4 – 1.6mm high-resolution ihMT and ihMTR. A volume reconstructed with half-sampling is also shown

Figure 3 – (a) ihMT-RAGE vs ihMT-FSE vs FLAIR-ihMT. Red arrows show area of improvement with FLAIR-ihMT compared to both ihMT-RAGE and FSE (b) comparison of ihMT image quality in infratentorial regions and (c) normalized ihMT signal comparison

Figure 2: Example magnetization transfer ratio (MTR), inhomogeneous MTR (ihMTR) and myelin water fraction (MWF) maps from five of the participants. ihMTR and MWF show more variation in white matter than MTR.

Figure 3: Pearson correlations between MTR and MWF and ihMTR and MWF. (ACR=Anterior corona radiata, AIC=Anterior limb of internal capsule, BOD=Body of corpus callosum, CEP=Cerebral peduncle, CIN=Cingulum, EXC=External capsule, GEN=Genu, PCR=Posterior corona radiata, PIC=Posterior limb of internal capsule, PTR=Posterior thalamic radiation, RIC=Retrolenticular part of internal capsule, SAS=Sagittal stratum, SCR=Superior corona radiata, SLF=Superior longitudinal fasciculus, SPL=Splenium)

Figure 2: The top panel consists of an overlay of GluCEST map for the entire slice (left), followed by the overlay of only hippocampal ROIs (middle) and the corresponding T1map of the slice from MP2RAGE (right) from a male (top) and two female volunteers (center, bottom).

Figure 1: GluCEST contrast values from the ROIs drawn on hippocampus of all the nine subjects as well as their rearranged values from hippocampus based on the side where GluCEST contrast was higher (termed as Lateral) and the other side (termed as Contralateral).

Fig.3 In vivo experiment of intranasal administration of 10%-PEG Iohexol Lipo (n=3). (A) Z-spectra, (B) CESTLDFit plots of CEST contrast at 4.3ppm, (C), T2 image of ICR mice brain, and (D), (E), (F), (G) CEST map Pre-Injection, and 0.5hr, 1hr, 1.5hr after intranasal injection at 4.3ppm.

Fig. 1 CEST properties of 1%-PEG Ioh-Lipo and 10%-PEG Ioh-Lipo (n=3). (A) Z-spectra and (B) MTRasym of Ioh-Lipo under 0.9 μT and pH 7.0. (C) was the B1 power optimization on 10%-PEG Ioh-Lipo.

Fig. 1. Conventional MR images, APTw images and masks for two post-treatment GBM patients. Top case, a patient with tumor recurrence; bottom case, a patient with treatment effects.

Fig. 2. CRT decision tree analysis on the basis of the data from FLAIR, APTw, and both.

Figure 3. [6,6’-²H]-glucose flux to lactate is localized to the tumor region in low-grade gliomas in vivo. Representative metabolic heatmaps from 2D CSI studies in mice bearing orthotopic BT257 tumor xenografts following injection of a bolus of [6,6’-²H]-glucose. Left panel shows a representative axial T2-weighted MR image while the right panel shows a metabolic heatmap of the SNR of lactate produced from [6,6’-²H]-glucose.

Figure 3: Overview of the possible 3D lesion localisation for multiple metabolites in patient #2, an FCD 1b patient. The hotspot location corresponds well to later focal resection.

Figure 5: Metabolic ratio maps overlayed to post-resection clinical imaging shows good correspondence of 7T metabolic imaging to surgical FCD resection.

Figure 3: A/ In vivo GABA edition using MEGA-PRESS pulse sequence. Examples of GABA and GLX peaks in the subtracted spectrum, fitted using the JET algorithm, in Sham (top) and KA-MTLE (bottom) mice (regular KA dose). B/ Regression curve between in vivo GABA+/GLX and ex vivo GABA/GLX

Figure 2: A/ Score plot of the principal component analysis (PCA) built with HRMAS MRS data of KA-MTLE mice and the 5 brain regions sampled (ipsilateral and contralateral hippocampus (IH and CH), posterior (P) and anterior (A), and adjacent cortex) . B/ Variable importance in the projection (VIP) ranking metabolites from most discriminative (VIP>1) to less discriminative ones in KA-MTLE mice C/ Percentage of variation of 8 metabolites (among the 19 metabolites quantified) in the 5 brain regions of KA-MTLE mice vs their sham homologous.

Figure 2: Illustrative axial views at three positions for all the retrieved parameter maps. From left to right: the bound pool fraction (F), the exchange rate (k_f), the transverse relaxation time of the bound pool protons (T_2r), the observable longitudinal relaxation time of the free pool protons (T_1,obs), and the B₁-map. Yellow and red ROIs shown on the middle row of the F slices were used for the data presented in Table 1.

Table 1: Mean and standard deviation of the bound pool fraction (F), the exchange rate (k_f), the transverse relaxation time of the bound pool protons (T_2r), the observable longitudinal relaxation time of the free pool protons (T_1,obs) calculated over ROIs selected in white matter (red rectangle) and gray matter (yellow rectangle) shown in the middle row of F slices in Figure 2.

Fig. 1. Averaged Z-spectra and AREX spectra for each pool (a, b), statistics of fitted AREX values for amide (c), NOE(-1.6) (d), and NOE(-3.5) (e) as well as S₀ (f) from the whole brain with a diffusion-weighting of b = 0s/mm² (blue) and 400s/mm² (red), respectively. Each AREX spectrum contains a CEST or NOE peak from a corresponding pool and a residual water peak. The great residual water peak is due to the use of the inverse analysis for calculating AREX values. * P < 0.05

Fig. 3. Averaged Z-spectra and AREX spectra for each pool (a, b), statistics of fitted AREX values for amide (c), NOE(-1.6) (d), and NOE(-3.5) (e) as well as S₀ (f) from the whole brain before (blue) and after (red) the injection of 5 mg/kg MION, respectively. * P < 0.05

Fig. 2: a) Representative anatomical image (T2-weighted) and superimposed colour-coded AACID maps for b) baseline and c) 60 minutes after injection of cariporide at day 14 post-implantation of the tumour. The tumour region is highlighted by the purple line. The average AACID value in tumour regions pre- and post-injection of cariporide was 1.22±0.02 and 1.27±0.020, respectively.

Fig. 1: The average AACID values for a) control groups (N=15) in right and left frontal lobes. b) experimental groups at 7-9 days (N=22) and 14-16 days (N=20) post tumour implantation. Error bars represent the standard error of the mean. The asterisks indicated p<0.05 in the paired t-test.

Figure 2. DWI, APTw, APT^#, and NOE^# images of a representative case using saturation B1 power of 1uT and 1.5uT.

Figure 5. DWI, APTw, APT^#, and NOE^# images of a representative case at a B1 power of 1uT.

Figure 2: Box and whisker plots with data points of MTR (magnetization transfer ratio), CSA (cross-sectional area), circularity, eccentricity, and nerve fascicle density by subject type. CMT1A is Charcot-Marie-Tooth Type 1A, CMT2A is CMT type 2A, and HNPP is hereditary neuropathy with liability to pressure palsy.

Figure 1 Left: Representative magnetization transfer (MT)-weighted MRI images of study subjects with segmented sciatic nerves in red in magnified inlay. Right: Magnetization Transfer Ratio (MTR) for subjects. Subjects in row A are healthy controls, B have Charcot-Marie-Tooth type 1A, C have CMT type 2A, and D have hereditary neuropathy with liability to pressure palsy (HNPP). Patient imaging values are shown in the table.

Figure 1. GluCEST maps: data from all subjects projected onto the anatomy of a single subject for visual representation. A) average gluCEST by segment, baseline (pre-stimulation) -15 subjects. B) post "sham" (placebo) stimulation - 5 subjects. C) post cTBS (real) stimulation -- 10 subjects. The colorscale is identical in all maps, with settings as shown in the screenshot from ITK-SNAP. Green arrow indicates the left precentral gyrus, the intended target of cTBS in stimulated subjects. Please see main text for full listing of anatomical segments treated distinctly in this analysis.

Figure 2. Barplot of gluCEST changes by segment: 99% confidence intervals (CI) of mean change, as report by unpaired T-test. A cluster of four bars is shown for each segment as listed in the text, beginning with the Precentral Gyrus for each side. A T-test was performed to estimate the 99% confidence interval for the change in each segment for both the sham and stimulated subjects. The 'true' change likely lies between the two confidence intervals, generally giving a value near zero for sham subjects, but a small negative value for stimulated subjects.

Figure 2: Illustration of a glucoCEST scan of a patient with a brain tumour under the influence of D-glucose infusion, rigid-head motion and dynamic lateral ventricle dilatation and contraction.

Figure 4: AUC (area under the curve) maps depicting different intervals: Pre-infusion (115 s), during infusion (245 s), and two post-infusion time intervals (45 s before, 160 s after peak) showing signal rise and decay. Five cases showing ground truth with D-glucose infusion and without motion (E), without infusion and with motion (A, C), and with infusion and motion (B, D), before (A, B) and after (C, D) retrospective motion correction.

Figure 4. Example gluCEST map through medial temporal lobe slice, shown in transparent overlay on the T₂-weighted structural image.

Figure 3. Comparison of gluCEST values in left and right hippocampal subfields. This boxplot is analogous to those in Figure 1, although rather than combining the corresponding measurements on the left and right sides of the brain, we inspect each separately. It can be seen from this plot that the gluCEST distribution in the dentate gyrus is higher than in the other subfields to a corresponding degree on both sides of the brain.

Figure 4. Coronal slices of the reconstructed T₁w images (A, E), mean MTR_asym maps within 2-4 ppm (B, F), T₁ maps (C, G), and T₂ maps (D, H) of a left TLE patient (first row) and a right TLE patient (second row). The MTR_asym values were higher in the epileptogenic hippocampus and amygdala (HA, illustrated by red boxes) than those in the contralateral HA. However, T₁ and T₂ values in the HA ipsilateral to the seizure were similar to or even lower than those in the contralateral HA (D, G, and H).

Figure 3. ROC curves for the lateralization of epileptic foci using metric1 (A) and metric2 (B). The two metrics used MTR_asym (blue line), T₁ (red line), and T₂ (yellow line) values as input indices. For both of these two metrics, AUC values of the MTR_asym index were significantly higher than those of T₁ and T₂.

Fig. 5: Representative images of control and shiverer mouse brains. T2-weighted image (leftmost) is compared with ihMTR image at a dual offset frequency of 12 kHz as well as MTR image at offset frequencies of 4, 8, 16 kHz. Corpus callosum (orange arrow), cerebral peduncle (yellow), and trigeminal nerve (green).

Fig. 3: Optimization of ihMTR signals in shiverer mouse brains (n=3). A series of dual offset frequencies ranging from 4k to 24k in horizontal axis; time delays from 0.01 ms to 4 ms in vertical axis; and saturation power in diagonal axis.

*Note: Using the one-way ANOVA and spearman’s correlation coefficient for ranked data

Figure 1. ROC curve of APT value and K^trans value in peritumoral edema between grade Ⅱ and Ⅲ (AUC: 0.971 for APT, 0.714 for K^trans; 95%CI: 0.89-1.00 for APT, 0.41-1.00 for K^trans).

Fig 4: Two representative cases for gliomas with aligned Contrast enhancement(A,D), APTw (B,E) and 3D-pCASL.A 48-year-old female with 1p/19q codeletion type (A,B,C),shows low APTw value and low perfusion. A 58-year-old male, with 1p/19q undeletion (E,F,G),shows high APTw value and high perfusion.

Fig 2: ROC curve for APTw and 3D-pCASL in distinguishing IDH mutation type and IDH wild type. ROC curve: receiver operating characteristic curve

Fig.1 Correlation between parameters on CEST imaging by multi pool model and MET T/N ratio. The Pearson's test showed correlation between MTRasymmean and MET T/N ratio (a) (r = 0.46), APT_T1mean and MET T/N ratio (b) (r = 0.53)

Fig.2 Comparison of average MTRasymmean (a), APT_T1mean (b) and MET T/N ratio (c) for IDH1-mutant glioma patients, IDH1 wildtype patients (* p<0.05).

Figure 3. Reconstructed and overlaid amide proton transfer-weighted (APTw) image of the whole-brain and the corpus callosum in an unsaturated image of a typical rat in each group (control group, normal control group; DEM, demyelinated group; and REM, remyelinated group). The black lines in the images of the whole-brain area for all groups indicate the contour of the corpus callosum. The color bar represents the range of calculated APTw signal (%).

Figure 2. Averaged Z-spectra (a), magnified Z-spectra from 0 to 6 ppm (b), magnified Z-spectra from -6 to 0 ppm (c) , magnetization transfer ratio asymmetry (MTRasym) curves (d), and amide proton transfer-weighted (APTw) signals calculated in the corpus callosum at each group (e) (Msat, signal intensity with RF saturation; M0, signal intensity without RF saturation; controls, black; DEM, demyelinated group, red; and REM, remyelinated group, green). **p = .009 and ***p < .001.

Figure 1. The schematics of CESTNet (A) and AREXNet (B).

Figure 3. Representative CEST and AREX maps of WT and AD brains, generated by trained CESTNet and AREXNet. CEST maps (A_APT, A_rNOE and A_MT) of central (A) and anterior slice (B). AREX maps (R_APT, R_rNOE and R_MT) of central slice (C) and anterior slice (D).

FIGURE 1 Schematic of data processing pipeline. Simulated background Z-spectrum (A) are generated. In each simulated background Z-spectrum (C), the data marked as red solid dots are inputted for training and the data in blue line are target for training. (D)The feedforward neural network. (B) The data marked as red solid dots from the acquired Z-spectrum are inputted for prediction. (E) The background Z-spectrum (marked as dashed blue curve) is obtained from the network. (F)The APT map and NOE map are obtained by subtracting background Z-spectrum from the acquired Z-spectrum.

FIGURE 3 The results from a patient with cerebral infarction. The 1st row includes the conventional T1-weighted image (A), T2-weighted FLAIR image (B), APT map (C) and NOE map (D). The 2nd row includes the fitting Z- spectrum of infarction tissue (E) and normal tissue (F), boxplots of APT contrast (G) and NOE contrast (H) between infarction and normal tissues. In (E) and (F), the red curves are B₀-corrected Z- spectra, and the blue lines are background Z- spectra from neural network fitting. The red arrow in the T2W FLAIR image indicates the area of the lesion.

Figure 5 – Tumor and Radiation Necrosis: (A) The median values with violin plots are shown for the tumor (red) and radiation necrosis (green) outcomes. Asterisks represent significant differences (**p<0.01, *p<0.05) between the two outcome groups after adjusting for multiple testing. (B) The ROC curve is shown of the significant parameter after multivariable modelling for predicting a tumor outcome.

Figure 3 – Example of Radiation Necrosis: The post-gadolinium T₁-weighted images are shown along with the four CEST maps – MTR amide and rNOE, acquired with B₁=0.52μT and B₁=2.0μT. The values on the bottom right represent the ROI medians and standard deviations.

Data are presented as median ± interquartile range unless otherwise noted.

Fig. 1. MTC magnitude images for an MSA-P patient, PD patient and HC are shown. From the MTC magnitude images of the MSA-P and PD patients, we can easily observe the NM depigmentation with lower contrast-to-noise than HC. The patients with MSA-P apparently suffer the worst NM degeneration.

Figure 1. 50-year old male with glioblastoma (Grade Ⅳ).

Glioblastoma (arrow) is shown in the left temporal cerebral suubcortical and white matter on Fluid-attenuated inversion recovery (FLAIR), Contrast-enhanced T1-weighted imaging (CE-T1WI) and CEST imaging. 3D CEST imaging can cover the whole lesion as compared with 2D CEST imaging. MTR_asym at 3.5ppm in this case were 2.12% (2D CEST imaging) and 1.69% (3D CEST imaging). 2D and 3D CEST imaging were diagnosed as glioblastoma (grade Ⅳ).

Figure 2. Results of correlation and Bland-Altman plot analysis between 2D and 3D CEST imaging measurement of MTR_asym (at 3.5 ppm) for all lesions.

3D CEST imaging had significant and excellent correlation with 2D CEST imaging (r=0.80, p<0.0001). The limits of agreement between 2D and 3D CEST imaging was -0.012±0.73 (mean±1.96 × standard deviation) %.

Figure 2. Pre-treatment axial APTw image of a primary nasopharyngeal carcinoma in a 44-year-old woman (APT_mean = 1.34%) and a 26-year-old woman (APT_mean = 2.64%).

Figure 1. The correlation between mean APT value (APT_mean) and percentage change (%Δ area) of primary tumor.

Figure 4. Representative metabolite maps (NAA, Cr, Cho) and spatially localized spectra obtained from a healthy subject using SPICE. High-quality spatiospectral distributions of various brain metabolites were obtained.

Figure 5. Simultaneously obtained water image, QSM map, T2* map at 1.0×1.0×3.0 mm³ nominal resolution and metabolite maps of NAA, Cr and Cho from two representative traverse images at 3.0×3.0×3.0 mm³ nominal resolution from a healthy subject in a single 8-min scan using SPICE.

Figure 1. T2-FLAIR image (a) of an FCD (white arrow) -associated epilepsy patient, the volumes of interest of HERMES in the patient (b and c) and in a matched healthy control (d). The mean (± standard deviation) GABA (e), Glx (e) and GSH (f) -edited spectra from the HERMES sequence in FCD foci, contralateral regions and healthy controls.

Figure 2. Comparation of GABA, GSH and Glx levels between the FCD foci, contralateral regions and healthy controls using ANOVA, and the results indicated that GABA levels was significantly increased in FCD foci compared with contralateral regions (p=0.007) and with healthy controls (p=0.003).

Figure 1. The coronal view high-resolution multimodal imaging of patient #3, including T1w-MPRAGE, PET-SUVR, NAA, Cr, Cho, Ins, NAA/Cr, NAA/(Cho+Cr), MWF, and QSM maps. Red arrows point to the ipsilateral temporal lobe.

Figure 4. Relationships between A) NAA/(Cho+Cr) and PET-SUVR in hippocampus; B) NAA/(Cho+Cr) and PET-SUVR in thalamus; C) NAA/(Cho+Cr) and MWF in cingulum hippocampus. P and r values are results of Pearson’s correlation.

Fig3. The group-wise directional pathways identified by Metabolic effective connection(MEC) method. Arrows represent the direction of the pathways. The line thickness represents the relative strength of the MEC. There were more bidirectional interactions between NAc and caudate, and the unidirectional pathway from OFC to caudate revealed the regulation in frontostriatal dopamine pathway. (One-sample t-test, p<0.05, fdr correted)

Fig2. The group-wise directional network identified by Granger causality. Arrows represent the direction of the pathways. The line thickness represents the relative strength of the granger causality index of that pathway. The SN and thalamus were almost isolated from other nucleus and cortex. Bidirectional connection were identified between the right OFC and left NAc. (One-sample Wilcoxcon signed rank test, p<0.05, fdr corrected)

Figure 1. High-resolution metabolite maps and PET images simultaneously acquired from a HC, an MCI patient, and an AD patient, respectively. A global NAA reduction and mI elevation were observed with the increased dementia severity.

Figure 3. Comparisons of neurometabolic concentrations among the HC, MCI, and AD groups for the global composite regions (top), PCC/precuneus (middle) and hippocampus (bottom). * p<0.05, ** p<0.01, *** p<0.001.

Figure 2 – Improvement of DGE ²H-MRS spectral quality upon MP-PCA denoising. GL7: (A) T2-RARE showing the tumor region selected for DGE ²H-MRS (yellow dashed line); (B) SNR_DHOi before and after denoising (average±SD); (C) left – eigenvalue spectrum from PCA decomposition (blue) and fit of MP distribution (orange), right – Gaussian distribution of denoising residuals, verified by the linearity of their logarithm; (D) DGE ²H-MRS data stacked, original and denoised, and spectral fitting including individual components, raw data, estimate and residual. * p=0.0003, paired t-Test.

Figure 3 – Robustness of the kinetic model for different tumors and improvement of Glx fitting with MP-PCA denoising. Tumor volumes selected for DGE ²H-MRS (yellow dashed line) overlaid on reference T2-RARE transversal images for each animal (GL5-9, top). Fitting of Glc (red line), Glx (green line) and Lac (blue line) time-course changes displayed, before (center) and after denoising (bottom).

Figure 2a - Sample volume of interest (65x15x20mm³­) in white matter region on MPRAGE ‘anatomical’ scan for healthy control.

b - Sample spectrum output from LCModel for each participant group with fit quality measures.

Figure 4 - Tissue content within VOI separated by subject group and measures of spectrum fit quality (Full-Width-Half-Maximum, FWHM and Signal-to-Noise Ratio, SNR); FWHM and SNR. Where WMf, GMf, CSFf and Lesionf, are fraction of white matter, grey matter, cerebrospinal fluid and lesional tissue, respectively.

Figure 4. Diagnostic accuracy of the Cho/Cr ratio in the differentiation between high grade and low grade gliomas.

Figure 1. Comparison of metabolite levels obtained from PRESS and edit-off spectrum using Bland-Altman analysis.

Figure 2. Representative in vivo ¹H MR spectra in the right hippocampus of cuprizone-treated rats. The figure shows the raw spectrum (gray), fitted spectrum (navy), baseline (light blue), and 14 individual metabolite fits below (brown).

Figure 3. Bar graph indicating the mean cerebral metabolite concentrations (a) and Cramer-Rao lower bounds (CRLBs) (b) in the right hippocampal region in control (CTRL) and cuprizone-treated (CPR) rats, quantified using the Linear Combination of Models software. The vertical lines on each of the bars indicate the (+) standard deviation of the mean values. *p < 0.05; **p < 0.01; ***p < 0.005.

Figure 1. Reconstructed MWF and neurometabolites maps of an acute mTBI patient and a healthy control. The neurometabolites include NAA, Cr, Cho and mI. The FOV covers the whole brain (240x240x120 mm³) and the data acquisition takes 8 minutes.

Figure 2. A reduction in MWF was observed in the occipital tract of acute mTBI patients group and the spatially resolved spectra revealed a reduction of NAA and increase of mI in the patient.

Figure 1. A demonstration of the MRS voxel in prefrontal cortex and spectrums of a blast and sham ferret at 3-day post-blast.

Figure 3. Glutamate and taurine levels in prefrontal cortex in sham and blast exposed ferret from 3-day to 3-month post-blast. *p<0.05, **p<0.01, group main effect; ##p<0.01, visit main effect.

Figure 3. Distributions of anterior cingulate cortex (ACC) a) phosphocholine + glycerophosphocholine (tCho) in the overall, riluzole, and placebo cohorts, and b) creatine + phosphocreatine (tCr) in the riluzole and placebo cohorts (* = p < 0.05, ** = p < 0.01).

Figure 4. Distributions of amygdala a) glutamine (Gln) in the overall and placebo cohorts, and b) glutamate + glutamine (Glx) in the riluzole cohort (* = p < 0.05).

Figure 1. A) The diffusion-weighted images of a stroke monkey brain demonstrated infarct evolution on Day 0, 2, and 4 post stroke (left). B) in vivo proton MR spectra in the contralateral and ipsilateral voxels in the stroke monkey brain 2 days post stroke. V1, MR Spectrum of contralateral voxel; V2, MR spectrum of ipsilateral area after MCA occlusion.

Figure 2. A) Illustration of the representative slices of a stroke monkey brain with regions of interest of contralateral secondary somatosensory cortex (S2)(S2-CON, seed for functional connectivity analysis) and the ipsilateral S2 (S2-Ipsi). B) representative correlation map of S2 in a stroke monkey brain before stroke surgery (Pre) and post surgery. p = 0.026 with 20 voxels as threshold.

Fig. 2. Distribution of the quantified metabolite ratios in the analyzed datasets. The plot shows the variability of tumor tissue samples in comparison to the normal-appearing tissue.

Fig. 4. Decision tree for the classification of tumor samples in normal-appearing tissue (normal) and tumor tissue.

Figure 4: Exemplary T1-, T2-weighted MR images and corresponding QSM (Quantitative Susceptibility Maps) of a LC patient (upper row) and a healthy control (lower row).

Figure 2: Exemplary spectra of two LC pacients and a healthy control.

Bar Charts of the distributions of GABA+ levels, normalized fitting errors, linewidth in Hz, NAA, Cr, Cho levels PD Parkinson’s disease, HC healthy control, NAA N-acetyl aspartate, Cr creatine, Cho choline

GABA+-edited spectra in the upper brainstem of all 36 participants, showing the intended signal at 3 ppm

Figure 4: Graphical representation of the GSH content in different brain regions. [A] Bar graphs of GSH/NAA at the thalamus and at the lateral ventricle with Student's t‐test result. It can be seen that there is a significantly less GSH/NAA content in the region around the lateral ventricle compared to the thalamus. [B] Bar graphs of GSH/NAA for the thalamus and lateral ventricular for each individual mouse. It can be seen that thalamus has significantly greater GSH/NAA than the lateral ventricle.

Figure 2: GSH detection in the phantom solution with J-edited 1H MRS. Figure indicates single-voxel spectra of the 30 mM GSH phosphate buffer solution (PBS) acquired at 37 ºC in 4 min. [A] ‘ON’ Spectra with editing pulse applied at 4.56 ppm. [B] ‘OFF’ spectra with editing pulse applied at 8 ppm. [C] Difference spectra between [A] and [B] showing edited GSH resonance at 2.98 ppm. [D] GSH peak graphs at the following TE: 68, 90, 110, 130 and 150. [E] Plot of the GSH peak area under the curve versus TE.

Figure 2. Bar graph showing coefficient of variations for different metabolites measurements made using different pulse sequences

Figure 1. Representative spectra showing spectra acquired from right basal ganglia using different pulse sequences.

Fig.3: Areas in GM that survived the Δ_%[NA] > 10% are shown for 2 randomly selected subjects. The arrows point to the "blob" that was identified under the marker on the MPRAGE (as demonstrated in Fig.1). The widespread pattern of Δ%[NA] was similar across the group.

Fig.1: The motor ROI was selected by overlapping the GM Δ_%[NA] image onto the subject's down-sampled MPRAGE. A MatLab script was written to identify "blobs" where Δ_%[NA] > 10%. The script yields a blob ID number which is then used to extract the motor ROI for further analyses. Note the spatial relationship between the marker and the ROI.

Figure 1. The first aMRI maps of an axial image slice of the awake, healthy resting human brain. Panel (a) is the T₁-weighted image. Panel (b) maps the cell density, r (cells/μL). Panel (c) maps the average cell volume, V (pL). Panel (d) maps the cellular water efflux rate constant, k_io (s^-1), reflecting cytoplasmic Sodium Pump enzymatic turnover. Note the particularly large k_io values in white matter [panel (d)].