tDKI-Net: A Joint q-t Space Learning Network for Diffusion-Time-Dependent Kurtosis Imaging.

Time-dependent diffusion magnetic resonance imaging (TDDMRI) is useful for the non-invasive characterization of tissue microstructure. These models require both densely sampled q-t space data for microstructural fitting, leading to very time-consuming acquisition protocols. To overcome this problem, we present a joint q-t space model-tDKI-Net to estimate diffusion-time dependent kurtosis and the transmembrane exchange, using downsampled q-t space data. The tDKI-Net is composed of several q-Encoders and a t-Encoder, designed based on the extragradient mechanism, each integrated with their respective mapping networks. In the tDKI-Net, two types of encoders along with their mapping networks are employed sequentially to generate kurtosis at individual diffusion times and to fit the transmembrane exchange time ( τm) using the time-dependent kurtosis according the Kärger's model. Meanwhile, we proposed a three-stage training strategy, including physics-informed self-supervised pretraining, DKI warm-up, and joint training, to match the network structure. Our results demonstrated that the proposed tDKI-Net could effectively accelerate tDKI scans, resulting in lower estimation error compared with other methods. Our proposed three-stage training strategy demonstrated superior results than those training from scratch, e.g., the normalized root mean square error (NRMSE) of τm decreased by up to 1.4%. We also investigated the training data size effects and found that although we used one-subject training, the network achieved lower NRMSEs for Kavg, K0 and τm (2.50%, 3.04%, 10.86%) than previous work that used three-subject training (3.8%, 9.5%, 12.1%). tDKI-Net can considerably reduce the scan time by 10.5- fold by joint downsampling the q-t space data without compromising the estimation accuracy.

Improving Microstructural Estimation in Time-Dependent Diffusion MRI With a Bayesian Method.

BACKGROUND: Accurately fitting diffusion-time-dependent diffusion MRI (td-dMRI) models poses challenges due to complex and nonlinear formulas, signal noise, and limited clinical data acquisition. PURPOSE: Introduce a Bayesian methodology to refine microstructural fitting within the IMPULSED (Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion) model and optimize the prior distribution within the Bayesian framework. STUDY TYPE: Retrospective. POPULATION: Involving 69 pediatric patients (median age 6 years, interquartile range [IQR] 3-9 years, 61% male) with 41 low-grade and 28 high-grade gliomas, of which 76.8% were identified within the brainstem or cerebellum. FIELD STRENGTH/SEQUENCE: 3 T, oscillating gradient spin-echo (OGSE) and pulsed gradient spin-echo (PGSE). ASSESSMENT: The Bayesian method's performance in fitting cell diameter ( d $$ d $$ ), intracellular volume fraction ( f in $$ {f}_{in} $$ ), and extracellular diffusion coefficient ( D ex $$ {D}_{ex} $$ ) was compared against the NLLS method, considering simulated and experimental data. The tumor region-of-interest (ROI) were manually delineated on the b0 images. The diagnostic performance in distinguishing high- and low-grade gliomas was assessed, and fitting accuracy was validated against H&E-stained pathology. STATISTICAL TESTS: T-test, receiver operating curve (ROC), area under the curve (AUC) and DeLong's test were conducted. Significance considered at P < 0.05. RESULTS: Bayesian methodology manifested increased accuracy with robust estimates in simulation (RMSE decreased by 29.6%, 40.9%, 13.6%, and STD decreased by 29.2%, 43.5%, and 24.0%, respectively for d $$ d $$ , f in $$ {f}_{in} $$ , and D ex $$ {D}_{ex} $$ compared to NLLS), indicating fewer outliers and reduced error. Diagnostic performance for tumor grade was similar in both methods, however, Bayesian method generated smoother microstructural maps (outliers ratio decreased by 45.3% ± 19.4%) and a marginal enhancement in correlation with H&E staining result (r = 0.721 for f in $$ {f}_{in} $$ compared to r = 0.698 using NLLS, P = 0.5764). DATA CONCLUSION: The proposed Bayesian method substantially enhances the accuracy and robustness of IMPULSED model estimation, suggesting its potential clinical utility in characterizing cellular microstructure. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 1.

Developing an AI-empowered head-only ultra-high-performance gradient MRI system for high spatiotemporal neuroimaging.

Recent advances in neuroscience requires high-resolution MRI to decipher the structural and functional details of the brain. Developing a high-performance gradient system is an ongoing effort in the field to facilitate high spatial and temporal encoding. Here, we proposed a head-only gradient system NeuroFrontier, dedicated for neuroimaging with an ultra-high gradient strength of 650 mT/m and 600 T/m/s. The proposed system features in 1) ultra-high power of 7MW achieved by running two gradient power amplifiers using a novel paralleling method; 2) a force/torque balanced gradient coil design with a two-step mechanical structure that allows high-efficiency and flexible optimization of the peripheral nerve stimulation; 3) a high-density integrated RF system that is miniaturized and customized for the head-only system; 4) an AI-empowered compressed sensing technique that enables ultra-fast acquisition of high-resolution images and AI-based acceleration in q-t space for diffusion MRI (dMRI); and 5) a prospective head motion correction technique that effectively corrects motion artifacts in real-time with 3D optical tracking. We demonstrated the potential advantages of the proposed system in imaging resolution, speed, and signal-to-noise ratio for 3D structural MRI (sMRI), functional MRI (fMRI) and dMRI in neuroscience applications of submillimeter layer-specific fMRI and dMRI. We also illustrated the unique strength of this system for dMRI-based microstructural mapping, e.g., enhanced lesion contrast at short diffusion-times or high b-values, and improved estimation accuracy for cellular microstructures using diffusion-time-dependent dMRI or for neurite microstructures using q-space approaches.

In vivo Mapping of Cellular Resolution Neuropathology in Brain Ischemia by Diffusion MRI.

Non-invasive mapping of cellular pathology can provide critical diagnostic and prognostic information. Recent developments in diffusion MRI have produced new tools for examining tissue microstructure at a level well below the imaging resolution. Here, we report the use of diffusion time ( t )-dependent diffusion kurtosis imaging ( t DKI) to simultaneously assess the morphology and transmembrane permeability of cells and their processes in the context of pathological changes in hypoxic-ischemic brain (HI) injury. Through Monte Carlo simulations and cell culture organoid imaging, we demonstrate feasibility in measuring effective size and permeability changes based on the peak and tail of t DKI curves. In a mouse model of HI, in vivo imaging at 11.7T detects a marked shift of the t DKI peak to longer t in brain edema, suggesting swelling and beading associated with the astrocytic processes and neuronal neurites. Furthermore, we observed a faster decrease of the t DKI tail in injured brain regions, reflecting increased membrane permeability that was associated with upregulated water exchange upon astrocyte activation at acute stage as well as necrosis with disrupted membrane integrity at subacute stage. Such information, unavailable with conventional diffusion MRI at a single t, can predict salvageable tissues. For a proof-of-concept, t DKI at 3T on an ischemic stroke patient suggested increased membrane permeability in the stroke region. This work therefore demonstrates the potential of t DKI for in vivo detection of the pathological changes in microstructural morphology and transmembrane permeability after ischemic injury using a clinically translatable protocol.

3D diffusion MRI with twin navigator-based GRASE and comparison with 2D EPI for tractography in the human brain.

PURPOSE: 3D pulse sequences enable high-resolution acquisition with a high SNR and ideal slice profiles, which, however, is particularly difficult for diffusion MRI (dMRI) due to the additional phase errors from diffusion encoding. METHODS: We proposed a twin navigator-based 3D diffusion-weighted gradient spin-echo (GRASE) sequence to correct the phase errors between shots and between odd and even spin echoes for human whole-brain acquisition. We then compared the SNR of 3D GRASE and 2D simultaneous multi-slice EPI within the same acquisition time. We further tested the performance of 2D versus 3D acquisition at equivalent SNR on fiber tracking and microstructural mapping, using the diffusion tensor and high-order fiber orientation density-based metrics. RESULTS: The proposed twin navigator approach removed multi-shot phase errors to some extent in the whole brain dMRI, and the 2D navigator performed better than the 1D navigator. Comparisons of SNR between the 2D simultaneous multi-slice EPI and 3D GRASE sequences demonstrated that the SNR of the GRASE sequence was 1.4-1.5-fold higher than the EPI sequence at an equivalent scan time. More importantly, we found a significantly higher fiber cross-section in the cerebrospinal tract, as well as richer subcortical fibers (U-fibers) using the 3D GRASE sequence compared to 2D EPI. CONCLUSION: The twin navigator-based 3D diffusion-weighted-GRASE sequence minimized the multishot phase error and effectively improved the SNR for whole-brain dMRI acquisition. We found differences in fiber tracking and microstructural mapping between 2D and 3D acquisitions, possibly due to the different slice profiles.

Diffusion-time dependent diffusion MRI: effect of diffusion-time on microstructural mapping and prediction of prognostic features in breast cancer.

OBJECTIVES: This study aimed to evaluate the effect of achievable td on the accuracy of microstructural mapping based on simulation and patient experiments, and investigate the feasibility of td-dMRI in distinguishing prognostic factors in breast cancer patients. METHODS: Simulation was performed using different td settings. Patients with breast cancer were enrolled prospectively between November 2020 and January 2021, who underwent oscillating and pulsed gradient encoded dMRI on a 3-T scanner using short-/long-td protocol with oscillating frequency up to 50/33 Hz. Data were fitted with a two-compartment model to estimate cell diameter (d), intracellular fraction (fin), and diffusivities. Estimated microstructural markers were used to differentiate immunohistochemical receptor status and the presence of lymph node (LN), which were correlated with histopathological measurements. RESULTS: Simulation results showed that d fitted from the short-td protocol significantly reduced estimation error than those from long-td (2.07 ± 1.51% versus 3.05 ± 1.92%, p < 0.0001) while the estimation error of fin was robust to different protocols. Among a total of 37 breast cancer patients, the estimated d was significantly higher in HER2-positive and LN-positive (p < 0.05) groups compared to their negative counterparts only using the short-td protocol. Histopathological validation in a subset of 6 patients with whole slide images showed the estimated d was highly correlated with measurements from H&E staining (r = 0.84, p = 0.03) only using the short-td protocol. CONCLUSIONS: The results indicated the necessity of short-td for accurate microstructural mapping in breast cancer. The current td-dMRI with a total acquisition time of 4.5 min showed its potential in the diagnosis of breast cancer. KEY POINTS: • Short td is important for accurate microstructural mapping in breast cancer using the td-dMRI technique, based on simulation and histological validation. • The 4.5-min td-dMRI protocol showed potential clinical value for breast cancer, given the difference in cell diameter between HER2/LN positive and negative groups.

Histological and molecular classifications of pediatric glioma with time-dependent diffusion MRI-based microstructural mapping.

BACKGROUND: Gliomas are the most common type of central nervous system tumors in children, and the combination of histological and molecular classification is essential for prognosis and treatment. Here, we proposed a newly developed microstructural mapping technique based on diffusion-time-dependent diffusion MRI td-dMRI theory to quantify tumor cell properties and tested these microstructural markers in identifying histological grade and molecular alteration of H3K27. METHODS: This prospective study included 69 pediatric glioma patients aged 6.14 ±â€…3.25 years old, who underwent td-dMRI with pulsed and oscillating gradient diffusion sequences on a 3T scanner. dMRI data acquired at varying tds were fitted into a 2-compartment microstructural model to obtain intracellular fraction (fin), cell diameter, cellularity, etc. Apparent diffusivity coefficient (ADC) and T1 and T2 relaxation times were also obtained. H&E stained histology was used to validate the estimated microstructural properties. RESULTS: For histological classification of low- and high-grade pediatric gliomas, the cellularity index achieved the highest area under the receiver-operating-curve (AUC) of 0.911 among all markers, while ADC, T1, and T2 showed AUCs of 0.906, 0.885, and 0.886. For molecular classification of H3K27-altered glioma in 39 midline glioma patients, cell diameter showed the highest discriminant power with an AUC of 0.918, and the combination of cell diameter and extracellular diffusivity further improved AUC to 0.929. The td-dMRI estimated fin correlated well with the histological ground truth with r = 0.7. CONCLUSIONS: The td-dMRI-based microstructural properties outperformed routine MRI measurements in diagnosing pediatric gliomas, and the different microstructural features showed complementary strength in histological and molecular classifications.

Time-Dependent Diffusion MRI for Quantitative Microstructural Mapping of Prostate Cancer.

Background Recently developed time-dependent diffusion MRI has potential in characterizing cellular tissue microstructures; however, its value in imaging prostate cancer (PCa) remains unknown. Purpose To investigate the feasibility of time-dependent diffusion MRI-based microstructural mapping for noninvasively characterizing cellular properties of PCa and for discriminating between clinically significant PCa and clinically insignificant disease. Materials and Methods Men with a clinical suspicion of PCa were enrolled prospectively between October 2019 and August 2020. Time-dependent diffusion MRI data were acquired with pulsed and oscillating gradient diffusion MRI sequences at an equivalent diffusion time of 7.5-30 msec on a 3.0-T scanner. Time-dependent diffusion MRI-based microstructural parameters, including cell diameter, intracellular volume fraction, cellularity, and diffusivities, were estimated with a two-compartment model. These were compared for different International Society of Urological Pathology grade groups (GGs), and their performance in discriminating clinically significant PCa (GG >1) from clinically insignificant disease (benign and GG 1) was determined with a linear discriminant analysis. The fitted microstructural parameters were validated by means of correlation with histopathologic measurements. Results In the 48 enrolled men, the time-dependent diffusion MRI measurements showed that higher GG was correlated with higher intracellular volume fraction and higher cellularity (intracellular volume fraction = 0.22, 0.36, 0.34, 0.37, and 0.40 in GGs 1-5, respectively; P < .001 at one-way analysis of variance), while lower cell diameter was found at higher GGs (diameter = 23.4, 18.3, 19.2, 17.9, and 18.5 µm in GGs 1-5, respectively; P = .002). Among all measurements derived from time-dependent diffusion MRI, cellularity achieved the highest diagnostic performance, with an accuracy of 92% (44 of 48 participants) and area under the receiver operating characteristic curve of 0.96 (95% CI: 0.87, 0.99) in discriminating clinically significant PCa from clinically insignificant disease. Microstructural mapping was supported by positive correlations between time-dependent diffusion MRI-based and pathologic examination-based intracellular volume fraction (r = 0.83; P < .001). Conclusion Time-dependent diffusion MRI-based microstructural mapping correlates with pathologic findings and demonstrates promise for characterizing prostate cancer. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Chatterjee and Oto in this issue.

Feasibility of oscillating and pulsed gradient diffusion MRI to assess neonatal hypoxia-ischemia on clinical systems.

Diffusion-time- (td) dependent diffusion MRI (dMRI) extends our ability to characterize brain microstructure by measuring dMRI signals at varying td. The use of oscillating gradient (OG) is essential for accessing short td but is technically challenging on clinical MRI systems. This study aims to investigate the clinical feasibility and value of td-dependent dMRI in neonatal hypoxic-ischemic encephalopathy (HIE). Eighteen HIE neonates and six normal term-born neonates were scanned on a 3 T scanner, with OG-dMRI at an oscillating frequency of 33 Hz (equivalent td ≈ 7.5 ms) and pulsed gradient (PG)-dMRI at a td of 82.8 ms and b-value of 700 s/mm2. The td-dependence, as quantified by the difference in apparent diffusivity coefficients between OG- and PG-dMRI (ΔADC), was observed in the normal neonatal brains, and the ΔADC was higher in the subcortical white matter than the deep grey matter. In HIE neonates with severe and moderate injury, ΔADC significantly increased in the basal ganglia (BG) compared to the controls (23.7% and 10.6%, respectively). In contrast, the conventional PG-ADC showed a 12.6% reduction only in the severe HIE group. White matter edema regions also demonstrated increased ΔADC, where PG-ADC did not show apparent changes. Our result demonstrated that td-dependent dMRI provided high sensitivity in detecting moderate-to-severe HIE.