Figure 1. A patient with IDH-wildtype GBM (Top) displays a CE lesion, which contains elevated ¹H perfusion. Accompanying HP-¹³C EPI data demonstrate the related elevation in total [1-¹³C]pyruvate and [1-¹³C]lactate signal. However, because of the perfusion abnormality, the k_PL map fails to provide visual conspicuity. By comparison, the patient with the non-enhancing IDH-mutant GBM (Bottom) displays a weakly perfused lesion, with lower relative signal from [1-¹³C]pyruvate and [1-¹³C]lactate. The k_PL map for this patient clearly highlights the lesion.

Figure 2. A ¹H MRSI CNI map overlaid on a T1w post-Gd image of a patient with IDH-wildtype GBM (A). Corresponding lactate-edited spectra from two adjacent voxels show CNI > 11 (green voxel) and lipid from the difference spectra (red voxel). Despite lower ¹H-lactate, the HP-¹³C data demonstrated abundant [1-¹³C]lactate production. Another patient with IDH-wildtype GBM presented with a larger infiltrative mass (CNI > 6; green voxel) as well as ¹H-lactate (B). In this case, HP [1-¹³C]lactate production was less pronounced, perhaps owing to the existing pool of steady-state ¹H-lactate.

Figure 5 – Glucose oxidation rate as a potential marker for GBM staging according to cell proliferation and stromal vascular fraction phenotype. H&E microphotographs showing how the expansile tumor growth progressively compresses and impairs vascular blood flow: I, small vessels, complete endothelial cell lining and sparse hemorrhages (GL6); II, blood flow obstructed, vasodilation and marked multifocal hemorrhages (GL7); III, necrosis with incomplete endothelial cell lining, vascular leakage, edematous stroma (GL5); IV, vascular depletion, edematous stroma (GL50).

Figure 1 – Experimental approach. A. in vitro, Seahorse XFp Mito-Stress Test with GL261 cells, measuring OCR and ECAR during sequential inhibition of the respiratory chain: oligomycin (ATP synthase), FCCP (H+ uncoupler), and antimycin A/rotenone (complexes I/III). B. in vivo, DGE ²H-MRS, showing the volume selection (pink), spectral fitting for each metabolite and the kinetic fitting; and DCE-T1 MRI, showing the tumor ROI (red), kinetic fitting and K^trans maps. C. post-mortem, Tumor histology for H&E and Ki67 – enlarged region showing Ki67 positive (red) and negative (blue) cells.

Figure 1. Heat map showing altered metabolites in the cell extracts of WMPY-1 and PCAFs incubated under normoxic and hypoxic conditions. Each column represents the mean value for at least 5 biological replicates. A two-tail unpaired t-test was used to compare values obtained under normoxic and hypoxic conditions in each cell line. * p ≤ 0.05, ** p < 0.01 and *** p < 0.001. MTA: GSH: glutathione; GSSG: oxidized glutathione.

Figure 3. Evolution of the mean invasion index (a) and the mean degradation index (b) relative to the initial time point over the 48 h of the experiment. * p<0.05 relative to normoxic conditions.

Figure 1: Representative individual dosimetry mapped in cGy onto the TBSS skeleton for an average-risk (A) and high-risk (B) patient.

Figure 5: TBSS results for interaction of dose with longitudinal evolution of the FA after irradiation. The skeleton is shown green overlaid on the average FA with red voxels demarcating significant negative interaction between dose and the positive change over time (slope) indicating that higher doses attenuated the rate of recovery. Distributions of these voxels were primarily in the cerebral peduncles, internal capsule, posterior thalamic radiation, corpus callosum, and corona radiata including the superior longitudinal fasciculus.

Schematic diagram illustrating pyruvate recycling pathway during [U-13C]glucose metabolism.

C2 lactate ¹³C NMR spectral region of brain tumor tissue extracts. The patients were given intravenous infusion of [U-¹³C]glucose prior to the surgical removal of the tumor. D12 and D23 doublets (via pyruvate recycling) represent [1,2-¹³C] and [2,3-¹³C]lactate isotopomers. The quartet (Q) signals correspond to the [U-¹³C]lactate that are generated from glycolysis of [U-¹³C]glucose.

Schematic representation of the study structure including A) data collection for MRI and tissue modalities and associated preprocessing, B) MR-histology coregistration, and C) mixed effect analysis structure

Mixed effect model results for cell density. The left-hand plots represent the first order models, whereas the right hand results represent the interaction between MR intensity and diagnosis, where blue = GBM and red = non-GBM.

This figure shows two representative IDHmt, non-CODEL astrocytomas that had two FLAIR images scanned with different TIs. Both cases highlight the importance of TI for FLAIR acquisition in terms of visualization of the T2-FLAIR mismatch sign.

The figures show the frequencies of presence or absence of the T2-FLAIR mismatch sign found in the TCIA/TCGA LrGG cohort. We divided the cohort into two groups according to the TI for FLAIR acquisition. The T2-FLAIR mismatch sign was more frequently positive for or IDHmt, non-CODEL astrocytomas if we acquired FLAIR with TI shorter than 2400 ms. The ROC curve for identifying IDHmt, non-CODEL astrocytomas is presented at the lower panel. The area under the curve improved from 0.63 to 0.87 87 if we acquired FLAIR with TI shorter than 2400 ms.

Figure 1: A) Group averages of corrected and uncorrected vessel size measurements within 4 ROIs: NAWM, NAGM, Tumor, CE-Tumor. Note that CE-Tumor ROI was delineated only for enhancing glioma patients (6 patients) , B) Scatter plots of corrected and uncorrected vessel size measurements , across subject within 4 ROIs (individual subject’s data are connected by solid lines). * represents significant p-value calculated from Student’s t-test (p < 0.05) .

Figure 2: Single slice of exemplary MRI images of two patients with enhancing gliomas ; A) Patient 01 and B) Patient 03. The images from left to right respectively are : T2W FLAIR; post contrast T1W; post contrast T1W with regions of interest, ROIs, overlaid (contrast-enhanced tumor (CE-Tumor) in yellow, tumor in orange, normal-appearing white matter (NAWM) in green, and normal-appearing grey matter (NAGM) in blue), calculated vessel size map (before application of leakage correction) and calculated vessel size map (after application of leakage correction).

Figure 1a 54-year-old female patient with meningiomas in brain: a. T2WI; b. T2WI FLAIR; c. T1-Gd; d. APTw image by SENSE-1.6 fused with T2WI FLAIR image; e-h. APTw images by CS-SENSE (with factors of 2, 3, 4, and 5) fused with T1-Gd image. The ROIs were obtained manually on the SENSE APTw image and copied to the others as shown.

Figure 2 a 64-year-old female patient with metastatic tumors in brain: a. T2WI; b. T2WI FLAIR; c. T1-Gd; d. APTw image by SENSE-1.6 fused with T2WI FLAIR image; e-h. APTw images by CS-SENSE (with factors of 2, 3, 4, and 5) fused with T1-Gd image. The ROIs were obtained manually on the SENSE APTw image and copied to the others as shown

Figure 1. Images of a 75-year-old female with an EGFR amplified grade III IDHwt glioma reveals T2 hyperintense tumor in the right frontal lobe, showing an infiltrative pattern (size of precontrast T1 abnormality much smaller than size of FLAIR abnormality). The histogram and cumulative histogram (dark blue line) of ADC shows a 5th percentile DC value and mean ADC value of 0.60 x 10-3 mm2/s and 1.01 x 10-3 mm2/s, respectively. The histogram and cumulative histogram (dark blue line) of nCBF shows a 95th percentile of the nCBF value of 7.54.

Figure 2. Images of a 58-year-old female with an EGFR non-amplified grade III IDHwt glioma reveals T2 hyperintense tumor in the left temporal lobe, showing an expansile pattern (size of precontrast T1 abnormality approximates size of FLAIR abnormality). The histogram and cumulative histogram (dark blue line) of ADC shows a 5th percentile ADC value and mean ADC value of 0.73 x 10-3 mm2/s and 1.28 x 10-3 mm2/s, respectively. The histogram and cumulative histogram (dark blue line) of nCBF shows a 95th percentile of the nCBF value of 4.55.

Fig.5: Upsampled HGG metabolic maps with size 184 x 184 using three measured patients data. From left to right (top), FLAIR, low-resolution (LR) HGG with size 46 x 46 after spectral quality (SQ) control, missing data filling in by inpainting (IPT) method and nonlocal means denoising (NLMD), super-resolution (SR) maps by Bicubic, weighted total variation (wTV), Unet and GAN. From left to right (bottom), LR HGG with size 46 x 46, LR HGG with size 46 x 46 after SQ control, missing data filling in by IPT method , SR maps by Bicubic+FNLM, wTV+FNLM, Unet+FNLM and GAN+FNLM.

Fig.4: Upsampled total N-acetylaspartate (tNAA) maps with size 184x184 using three measured healthy subjects data. From left to right (top), high-resolution (HR) MEMPRAGE (MPRG), low-resolution (LR) tNAA with size 46 x 46 after spectral quality (SQ) control, missing data filling in by inpainting (IPT) method and nonlocal means denoising (NLMD), super-resolution (SR) maps by shown methods. From left to right (bottom), LR tNAA with size 46 x 46, LR tNAA with size 46 x 46 after SQ control, missing data filling in by IPT method , SR maps by shown methods.

Fig. 1. A) T₁RESS pulse sequence using balanced (top) and unbalanced (bottom) steady-state readouts. A non-spatially selective contrast-modifying RF pulse (CMα) is applied periodically over the entire duration of the echo train to introduce an arbitrary amount of T1 weighting. B) Phantom consisting of serial dilutions of gadobutrol imaged with bT₁RESS using CMα values of 0^o, 30^o and 60^o. Note that there is negligible T1 contrast for a CMα flip angle of 0^o, but substantial T1 contrast is apparent as the flip angle is increased to 60^o.

Fig. 2. Pulse sequence comparisons after contrast administration in a patient with metastatic melanoma. 10-mm thick axial (top) and coronal (bottom) maximum intensity projections are shown to highlight differences in lesion visibility for uT₁RESS (left) and 3D spoiled GRE (right). Small metastatic lesions are much better visualized using uT₁RESS than with 3D spoiled GRE. The combination of twofold increased tumor-to-background contrast, improved CNR and suppression of intravascular signals with uT₁RESS is helpful to unambiguously identify small metastases.

Multivariable logistic regression analysis (including age, rADC and selected imaging features) was used to predict IDH status. (A) ROC curves of the multivariable probabilities for model 1 and model 2 show similar model performance in the study set. Model 1 consisted of rADC, age, enhancement pattern, calcification, cystic change and hemorrhage. Model 2 consisted of rADC, age, enhancement pattern, cystic change, hemorrhage and absence of calcification. (B) ROC curves of the multivariable probabilities for model 1 and model 2 show similar performance with the test set.

Comparison of AUCs among machine learning models. Receiver operating characteristic curves shown for logistic regression (Log Reg), support vector machine (SVM), Naive Bayes (NB) and Ensemble (random forest + eXtreme Gradient Boosting) in predicting IDH status of glioma. Log Reg yielded the greatest AUC for single-model prediction.

DCE MRI of a grade IV glioblastoma (figure a). (b) and (c) Ktrans and Ve using feeding artery of the tumor, respectively. (d) sampled AIF. (e) QTM velocity magnitude map. (f) and (g) Ktrans and Ve using carotid artery of the tumor, respectively. (h) sampled AIF.

DCE MRI of a grade II glioblastoma (figure a). (b) and (c) Ktrans and Ve using feeding artery of the tumor, respectively. (d) sampled AIF. (e) QTM velocity magnitude map. (f) and (g) Ktrans and Ve using carotid artery of the tumor, respectively. (h) sampled AIF.

FIG 4. (A,B) ROC analyses of the 2HG estimates from 43 patients are presented for Fitting methods 1 and 2. A red circle on an ROC curve corresponds to the smallest distance to the upper-left corner of the curve, at which a cutoff value was obtained as 1.3 and 0.8 mM for Fitting methods 1 and 2, respectively. (C,D) 2HG estimates from 24 IDH mutated and 19 IDH wildtype tumor patients are shown for Fitting methods 1 and 2. Green lines indicate the cutoff values from the ROC curves. For Fitting method 2, the 2HG estimates in IDH mutated tumors were 0.8 - 6.8 mM and those in IDH wildtype tumors were 0 - 0.4 mM.

FIG 1. Representative in vivo PRESS TE 97 ms spectra from three glioma patients are presented with LCModel fitting outputs and voxel positioning on T2-FLAIR images. LCModel-returned 2HG signals are shown with 2HG estimates and CRLB in brackets. The LCModel built-in lipid basis set (Fitting method-1) included Lip09, Lip20, and Lip13. New lipid basis set (Fitting method-2) included LipNew1, LipNew2, and Lip13. Dotted lines denote exclusion of the lipid signals in the basis set. Spectra are scaled with respect to the water signal from the voxel. Vertical lines are drawn at 2.25 and 0.9 ppm.

Figure 2: Typical plasma contrast-agent concentration curves C_p(t) extracted from the horizontal segment of the middle cerebral artery (arrow head), MCA (A); vertical segment of the ICA (short arrow), carotid syphon (B); and vertical segment of the SSS (long arrow) (C) following a bolus injection of 0.023 mmol/kg of GBCA. The 1^st-pass data are fitted using a gamma variate function, which excludes contrast-agent bolus recirculation. 3D T1-weighted gradient recalled echo mages obtained from a patient with a sporadic VS scanned at 1.5T.

Figure 3: Relationship between extracted VIF features and administered GBCA dose for VIF_SSS, VIF_MCA and VIF_ICA. Top row: Scatter plots of VIF peak (mM/L) vs administered GBCA dose (mM/kg). Bottom row: Scatter plots of area (= peak∙FWHM: full-width at half-maximum, mM·s /L) of VIF vs administered GBCA dose.

Figure 1. Global graph theory metrics as a function of dose (EQD2) in Gy. A. Global efficiency and B. Transitivity reported as the percent change from baseline to 6 months. Solid lines are linear fits and shaded areas represent the 95% confidence interval. Pearson’s Correlation Coefficient

Figure 2. Local graph theory metrics for subcortical regions. Percent change in A. Local efficiency and B. Clustering Coefficient as a function of EQD2 (Dose) in Gy. Solid lines are linear fits and shaded areas are the 95% confidence interval.

Figure 1. Representative images of MPRAGE (A, D), PETRA (B, E) and DANTE-SPACE (C, F) of a 57-year-old male with cerebral metastases of lung cancer. A-C: a well-enhanced lesion (arrowheads) near the cerebral falx in the right frontal lobe is seen in all sequences. The lesion is more conspicuous with DANTE-SPACE due to blood vessel suppression and high CNR, and could have been mislabeled as a small vessel with MPRAGE and PETRA. D-F: a small homogeneously meningioma in the left temporal lobe was displayed clearly in all sequences with clear margin

Figure 2. Post-contrast T1-weighted images of an enhancing tumor in right temporal lobe in a 37-year-old female. The signal intensity within this tumor was relatively homogeneous in MPRAGE (A) and PETRA (B), but it was heterogeneous in DANTE-SPACE (C). The enhancing effect (arrows) of adjacent meninges is more continuous in PETRA. The tumor signal from DANTE-SPACE (C) closely resembles pre-contrast T1WI and T2WI (G, H), and it was identified as a myopericytoma by surgical pathology (I).

Figure 2. Representative images showing a comparison of post-contrast Wave-CAIPI T1 MPRAGE and standard T1 MPRAGE sequences. (A) A 56-year-old male with brain metastases from squamous-cell carcinoma. (B) A 71-year-old male with brain metastases from non-small cell lung cancer.

Figure 1. Balloon plot showing the results of the head-to-head comparison of post-contrast standard T1 MPRAGE and post-contrast Wave-CAIPI T1 MPRAGE for visualization of abnormal intracranial enhancing lesions, the perception of noise, presence of artifacts due to motion, and the overall diagnostic quality. A zero-score indicates equivalency, negative scores (left) favor standard T1 MPRAGE, and positive scores (right) favor Wave-T1 MPRAGE.

Multimodalities integration in a 60-year-old male with left frontal bAVM in DiffusionGo. S1 and S2 Speech arrest was defined as discontinuation in number counting without simultaneous motor response by DCS (a). AF (red), SLF II (green), and posterior segment of SLF (yellow) were automatically reconstructed (d). Comparing with language BOLD functional activation (c), Tract-based cortical termination (AF termination in red; SLF II termination in green; orange color represented the overlap of AF and SLF II) has more sensitivity to manifest with DCS results (e).

A 41-year-old female with temporal lobe glioblastoma was reconstructed and displayed in DiffusionGo. Cortical surface (white), blood vessel (gold), tumor (green), and peritumoral edema (light blue with translucent) were integrated (a). The lateral view and superior view of the tumor, dorsal language pathway (AF, red; SLF II, green), and ventral language pathway (IFL, light blue; IFOF, light green; UF, pink). The termination projection of AF was shown in d. Superior displacement of left Wernicke’s area was identified with intact AF projecting to left premotor and left Broca areas.

ROC curves of DKI and DSC MRI parameters for differentiating glioma recurrence from pseudoprogression.

A 21-year-old man with tumor recurrence confirmed by repeat surgery was initially a high-grade glioma (WHO IV). An increasing enhanced lesion in the right frontal and temporal lobes was detected on CE-T1WI with increased CBV, CBF, MTT, TTP, MK, Ka and Kr values, except for MD and FA values.

Figure 1: Distortion-free EPI-MRE magnitude, NOSS map, and MP-RAGE image from a pituitary adenoma patient.

ROC curves showed that the AUC of uniformity (0.818) and correlation (0.711) in contrasted T1WI were greater than the other features among three sequences, so uniformity and correlation showed better diagnostic efficiency for predicting the proliferation status glioma.

Figure 1. MD (A), MK (B) and FA color (C) maps in a 35-year-old man with IDH 1-mutated glioma (WHO II); and MD (E), MK (F) and FA color (G) maps in a 36-year-old woman with IDH 1 wild-type glioma (WHO II); ROIs of glioma foci, pWM and cNAWM were drawn on non-smoothed MD maps (D and H). MD = mean diffusion; MK = mean kurtosis; FA = fractional anisotropy; pWM = perilesional white matter; cNAWM = contralateral normal appearing white matter; ROI = region of interest.

Figure 3. ROC curves of MK and MD values calculated from the glioma foci for differentiating IDH 1 mutant and wild-type low-grade gliomas. The AUC of MK was 0.88 (95% CI, 0.74~1.00), MD was 0.86 (95%CI, 0.73~0.99). MK = mean kurtosis; MD = mean diffusion.

PET-MR of a patient with left WHO IV glioblastoma treated by surgery and adjuvant radio-chemotherapy.

Upper row: A: enhancement on the frontal lesion on a contrast-enhanced axial T1-w image; B: hyper-perfusion with increased CBF on ASL images corresponds to the area of contrast enhancement; C: 18F-DOPA PET map with increased SUV

Lower row: SPM maps were concordant with significant clusters, which confirmed progression. The red ASL clusters appear to be more peripheral than the blue extensive PET clusters. D: T-score map; E: Z-score of asymmetry index; F: isocontour map.

PET-MR of patient with right glioblastoma, treated by surgery and adjuvant radiotherapy

Upper row: A: linear cavitary enhancement observed on the initially-treated frontal lesion site on a contrast-enhanced axial T1-w image; B: the absence of hyper-perfusion as assessed by CBF on ASL images; C: 18F-DOPA PET map showing no significantly increased SUV

Lower row: SPM maps were concordant, with the absence of significant clusters, which confirmed pseudo-progression. D: T-score map; E: Z-score of asymmetry index; F: isocontour map.

Figure 1. Evaluation of the spectra based on quality and reliability of the 2HG fit. Examples of each group have been shown in (B).

Figure 3. The relationship between 2HG concentration and the concentration of Glu+Gln and GABA in ‘Questionable’ (A) and ‘Reliable’ (B) fits. ×: information on voxel location in the tumor tissue is not available. Small marker size: no tumor (likelihood of 0-25%); big marker size: tumor (likelihood of 50-100%).

Fig. 3 T1 anatomy MRI; (from left) Pre-surgical @3T, Post contrast at1.5T (MPRAGE, dual T1 MPRAGEs) and AADC infusion at1.5T (MPRAGE, dual T1 MPRAGEs). Note that new dual T1 MPRAGE at1.5T can provide higher contrast of Put than that at 3T. In addition, dual T1 MPRAGE provides good whole brain anatomy. White spots are the targeted infusion regions.

Fig. 2 3D MPRAFGE MR images at effective TI 330 (A and A’), 350 (B), 370 (C) and 880 msec (D) at kz = 0; only axial images are shown. The images in A and A’ were acquired at a week apart sessions. Black arrows indicate the aliasing artifact in conventional MPRAGE image. Note that the signal in the white mater at optimal TI ~330 msec is well suppressed and the contrast of putamen (white arrowhead) and glubus pallidus (black arrowhead) vs. white mater is much higher than those of conventional MPRAGE image at effective TI 880 msec. In addition, the optimal TI MPRAGE MRI is reproducible.

LCModel analysis of an IDH mutant glioma patient's spectrum with and without cystathionine in the basis set

Metabolite ratios of cystathionine over total choline in codeleted and no-codeleted gliomas

Figure1 Different tumor information was shown between contrast-enhanced T1W images and 3D-ASL images. A: contrast-enhanced T1W images; B:3D-ASL; C: 3D-ASL fusion registration to contrast-enhanced T1W images.According to the enhanced area shown in the figure A, figure B showed the uneven distribution of CBF in tumor. The high cerebral blood flow area was mainly located on the right side of the enhanced edge, while the low cerebral blood flow area and the enhanced area overlap in a large region. Moreover, the fusion image C showed this result more clearly.

Table 1 The Characteristics of 50 single BM patients

Representative multi-parametric maps of two rat brains with an RN in the left hemisphere (top) and a 9L tumor in the right hemisphere (bottom).

Summarized ROI-based MRI parameters in the differentiation of contralateral normal appearing brain tissue (n=9), radiation necrosis (RN, n=5), and 9L brain tumor (n=4).

Figure 4. Non-contrast FLAIR images with T2Prep 400ms and TI 2100ms of a representative healthy subject. a, b Axial slice through the inferior part of the vestibule. c, d Sagittal reference slice. The saccule and utricle are separately visualized. Delineation of two structures is very similar to what is observed with contrast-enhanced FLAIR images in the literature (Attyé et al³). PSC posterior semicircular canal, LSC lateral semicircular canal, SSC superior semicircular canal, white arrow cochlea, dotted white arrow vestibule.

Figure 3. Peri- and endo-lymphatic space signal curves with variable TIs (224-5000ms) and T2Prep (0-600ms). Error bars indicate standard errors. a, b Red lines show the differences of the null points, which is larger when using 400ms T2Prep than 200ms. Green lines show the signal difference of peri- and endolymphatic spaces at endolymphatic space’s null point. The relative contrast of the two space was larger when using 400ms T2Prep than 200ms.

Figure 2: Example images. Methods 1, 2 and 3 are shown in (a), (b) and (c) respectively. An example of a segmentation mask is shown in (d): the main trunk is segmented in red, the upper division in green and the lower division in blue.

Figure 1: MRI sequences and acquisition parameters.

Figure2 : APTw values of the cervical malignant tumor group were significantly higher than those of the benign tumor group

Figure3: The ROC curve of the APTw value to differentiate between the malignant and benign neck tumors.

Figure 1: Representative cross-sectional 3D bSSFP images for a patient at 0.5T (a) and 3T (b) reformatted into various labeled planes (i.-iv.). Key anatomical structures being rated by 2 board certified radiologists are shown. Blue arrow: superior semicircular canal; Yellow arrow: facial (top) and cochear (bottom) nerve; White arrow: superior vestibular (top) and inferior vestibular (bottom) nerve; Green arrow: vestibule; Red arrow: cochlea.

Table 1: Statistical comparison of 0.5T vs clinical field strength scans

Images showing the acquired b0 (A), b1000 (B) and apparent diffusion coefficient (C) images with the region of interest for brain tumour as well as the noisy (D) and noise-suppressed (E) proton magnetic resonance spectroscopy of a patient with a medulloblastoma.

Boxplots comparing the balanced classification accuracy estimated through leave-one-out (A-B) or six-fold (C-D) cross validation and determined through linear discriminant analysis (A, C) or support vector machine (B, D) from diffusion weighted imaging (DWI) only, noise-suppressed proton magnetic resonance spectroscopy (NS-MRS) only, or combining NS-MRS and DWI.

Levels of significance: *, P<.05; **, P<.01; ***, P<.001; ****, P<.0001.

Fig 3. The color map of the patient (a, b) shows a green color of the base of tongue tumor and right lymph node metastases suggestive of low ADC value with non-zero ΔR2*, oxygenated tumor. The color map of another patient (c, d) of right tonsillar carcinoma shows pink-magenta color posteriorly, suggestive of near-zero ΔR2* and high PET activity, features of hypoxic tumor. This patient had early recurrence and disease-related mortality

Fig 2. 2D scatter histograms comparing (a) ADC to DR2*, (b) FDG-PET to ΔR2*, and (c) FDG-PET to ADC. A large standard deviation (length of vertical and horizontal bars) and ΔR2* are observed in a patient with recurrence (subject 6).

Figure 1. T2-weighted imaging (T2WI), b0 images of 2D TGSE-BLADE-DWI, and RESOLVE-DWI were fused. Color- and gray-coated images were derived from the T2WI and the two DWI b0 images, respectively. The T2WI and TGSE-BLADE-DWI b0 images matched well, showing the left orbital tumor (long arrows), ethmoid and maxillary sinuses (short arrows). The T2WI and RESOLVE-DWI b0 images were mismatched due to geometric distortions.

Figure 2. A 50-year-old male with a left orbital tumor. Compared with the T2-weighted imaging (T2WI) images and contrast-enhanced T1WI images, this lesion could be visualized clearly with TGSE-BLADE-DWI; no geometric distortions were seen. Slight distortions were seen with RESOLVE-DWI. More geometric distortions and ghosting artifacts were observed in the frontal lobe and ethmoid sinus with RESOLVE-DWI.

Figure 1. A 50-year-old female with a left internal auditory canal acoustic neuroma (red arrows). The lesion is clearly visible on the TGSE-BLADE-DWI scans without geometric distortion, but with moderate and severe distortions on the RESOLVE-DWI and SS-EPI-DWI scans, respectively, interfering with lesion detection. Ghosting artifacts (white arrows) were absent at bone-air interfaces and anatomic structural identifications (blue arrows) were superior on the TGSE-BLADE-DWI scans compared with the other two DWI methods.

Table 3. A comparison of the subjective evaluation results among the three diffusion-weighted imaging methods.

Figure 1.The IRIS sequence chart. IRIS adopts multi shot acquisition which signal collection into k-space is divided in the phase direction. Multi shot acquisition tends to cause ghost artifacts, so IRIS applies navigator-echo for each shot to make phase corrections.

Figure 2. DWI of the optic nerve on SSh EPI-DWI, SSh TSE-DWI, MSh. TSE-DWI, and IRIS. IRIS shows the least artifacts and image distortion, and the optic nerve was demonstrated with the highest image contrast on IRIS.

Figure1. 21-year-old male with lymphatic malformation (L to R: cranial to caudal level).

T2-weighted images obtained with CS with PI and with conventional PI clearly demonstrate the multiple cystic lesions extending from parotid space to parapharyngeal space (arrows). Neither image showed significant artifacts and both had the same overall image quality. The examination time of CS with PI (76 sec) was much shorter than that of conventional PI (126 sec).

Figure 2. Results of comparison for each quantitative image quality index among CS with PI and conventional PI.

SNR of CS with PI (11.2±3.6, mean ± standard deviation) was significantly higher than that of conventional PI (8.9±2.6, p<0.0001). %CV of CS with PI (9.6±3.0) was significantly lower than that of conventional PI (11.9±3.5, p<0.0001). CNR of CS with PI (7.7±2.9) was significantly higher than that of conventional PI (6.1±2.2, p<0.0001).

Figure 1 A ROI segmentation in a randomly selected case based on sagittal transverse T2WI. (A) Images from a representative PTC patient without ETE (female, 47 years old). (B) Images from a representative PTC patient with ETE (female, 37 years old).

Table 3 Multivariate logistic regression analysis results

A 43ys patient presenting with SSNHL and tinnitus of the left ear for 7 days. The PTA was 80 dB. 3D-FLAIR images was scanned 3 days later, showed that SIs of the left cochlea and vestibule were slightly higher than those of the right with markedly delayed enhancement after contrast. Other sequences showed no abnormality. This case was diagnosed as unilateral SSNHL due to labyrinthine inflammation combined with blood labyrinth barrier disruption. The patient was hospitalized and received a comprehensive therapeutic protocol for 9 days, the curative effect was ineffective as expected.

ROC curves of the two primary clinical-image prognosis assessment models for unilateral SSNHL.

Fig a. Model used for no hearing recovery with a moderate prediction value. It included six prognostic factors: age ≥ 50 years, simultaneously abnormal DPOAE and TEOAE, longer period from onset to clinical examination, PTA at onset, severe to profound initial hearing loss, and positive MR inner ear findings. Fig b. Model used for complete hearing recovery with a high prediction value. It consisted of two prognostic factors: PTA at onset and the period from onset to MR examination.

Fig. 3. Volunteer images acquired using the proposed SPLICE-PROPELLER protocols demonstrate good image quality without distortion artifacts.

Fig. 4. Patient results of the pituitary gland with the proposed SPLICE-PROPELLER protocol. Clear delineation of the pituitary gland with consistent quality is achieved in the clinical settings.

Figure 1. The first row, from left to right: Group B EPI, SENSE2.6, CS2.6-CS7.6. The second row, from left to right: Group C EPI, SENSE2.6, CS2.6-CS7.6

Figure2.ROIs for measurements of ADC. The measured left and right ADC was 0.81×10-3 and 0.69×10-3, respectively. The ADC value of the lesion is 0.47×10-3.

Fig. 3 A ROC curve was made based on the six most statistically significant parameters for differentiating glioblastoma and CNS lymphoma.

An ROC curve was made based on the six most statistically significant parameters for differentiating glioblastoma and metastatic tumors.

An ROC curve was made based on the six most statistically significant parameters for differentiating CNS lymphoma and metastatic tumors.

Fig 2.The gray histogram of CNS lymphoma、Glioblastoma and metastatic tumors. The abscissa represents different gray-scale values in ROI, and the ordinate represents the frequency of occurrence of gray-scale values.

Figure 2 – Example time series of swallow. Selected images from a healthy volunteer (every fifth slice) are shown from left to right, starting with the top row. In the middle slices, the pharynx constricts and the hyoid and pre-epiglottic fat elevate.

Figure 3 – Deformation maps and intensity change maps. Deformation magnitude is shown in (a) and is largest in the region of the tongue base and larynx. Intensity changes are shown in (b); in this case there are small intensity changes in the pharynx due to unsaturated tissue being pulled into the imaging slice.

Figure 3. Kaplan-Meier curves of LRRFS (a), LRFS (b), PFS (c) for NPC patients stratified as the K_^trans-low and K_^trans-high groups; and LRRFS (d), LRFS (e) for NPC patients stratified as the f_pv-low and f_pv-high groups; and DMFS (e) for NPC patients stratified as the v_e-low and v_e-high groups. K^trans_-low/f_pv-low/V_e-low group = patients with a primary lesion pretreatment K^trans/f_pv/V_e value ≤ 0.353/min, 0.034 and 0.289, respectively; K^trans_-high/f_pv-high/V_e-high group = patients with a primary lesion pretreatment K^trans/f_pv/V_e value > 0.353/min, 0.034 and 0.289, respectively.

Figure 4. Kaplan-Meier curves of LRRFS (a), LRFS (b), PFS (c) and DMFS (d) for NPC patients stratified as the v_{e-perc.90%-low} and v_{e-perc.90%-high} groups; and LRRFS (e), LRFS (f) for NPC patients stratified as the K_{ep-perc.10%-low} and K_{ep-perc.10%-high} groups. v_{e-perc.90%-low}/K_{ep-perc.10%-low} group = patients with a primary lesion pretreatment v_e-perc.90%/K_ep-perc.10% value ≤ 160, 75.5, respectively; v_{e-perc.90%-high}/K_{ep-perc.10%-high} group = patients with a primary lesion pretreatment v_e-perc.90%/K_ep-perc.10% value > 160, 75.5, respectively.

Figure 2: When starting the movement from a protrusive mandibular position, the sliding window reconstruction with a frame rate of 125 ms allowed the visualization of the repositioning of the condyle (dotted arrow) on the articular disc (solid arrow) (B-E) with subsequent full opening of the mandible (G). During the closing motion the condyle can once again be observed to slip down from the articular disc (J-L).

Figure 3: Visualization of the jaw motion with an incorporated occlusal splint. Starting in a therapeutic, more anterior and caudal initial position with the condyle (dotted arrow) already positioned on the articular disc (solid arrow), no additional repositioning or disc displacement was observed during the entire motion of the jaw.

Descriptive measures of the predicted IDH status from the visual inspection

Information provided to participants at the beginning of the survey.

Figure 3: Comparison of SNR and root canal visibility between the 32ch head coil (left), LC4 (center) and IFC coupled inductively to LC4 (right).

Figure 2: Ex-vivo measurements of a phantom show SNR gain of 6 with the inductively-coupled coil (IFC) in comparison to solely using the 4cm loop coil by Siemens (LC4).