Accurate actual flip angle imaging (AFI) in the presence of fat.

PURPOSE: To investigate the effect of incomplete fat spoiling on the accuracy of B1 mapping with actual flip angle imaging (AFI) and to propose a method to minimize the errors using the chemical shift properties of fat. THEORY AND METHODS: Diffusion-based dephasing is the main spoiling mechanism exploited in AFI. However, a very low diffusion in fat may make the spoiling insufficient, leading to ghosts in the B1 maps. As the errors retain the chemical-shift signature of fat, their impact can be minimized using chemical-shift-based fat signal removal from AFI acquisition modified to include multi-echo readout. The source of the errors and the proposed correction were studied in simulations and phantom and in-vivo imaging experiments. RESULTS: Our results support that AFI artifacts are caused by the incomplete fat spoiling present in clinically attractive short TR acquisition regimes. The correction eliminated the ghosting and significantly improved the B1 mapping accuracy as well as the accuracy of R1 mapping performed with AFI-derived B1 maps. CONCLUSIONS: The incomplete fat signal spoiling may be a source of AFI B1 mapping errors, especially in subjects with high fat content. Achieving complete fat spoiling requires longer TR, which is undesirable in clinical applications. The proposed approach based on fat signal removal can reduce errors without significant prolongation of the AFI pulse sequence. We propose that, when attaining complete fat spoiling is not feasible, AFI mapping should be performed in a multi-echo regime with appropriate fat separation or suppression to minimize these errors.

B1 inhomogeneity-corrected T1 mapping and quantitative magnetization transfer imaging via simultaneously estimating Bloch-Siegert shift and magnetization transfer effects.

PURPOSE: To introduce a method of inducing Bloch-Siegert shift and magnetization Transfer Simultaneously (BTS) and demonstrate its utilization for measuring binary spin-bath model parameters free pool spin-lattice relaxation ( T 1 F $$ {T}_1^{\mathrm{F}} $$ ), macromolecular fraction ( f $$ f $$ ), magnetization exchange rate ( k F $$ {k}_{\mathrm{F}} $$ ) and local transmit field ( B 1 + $$ {B}_1^{+} $$ ). THEORY AND METHODS: Bloch-Siegert shift and magnetization transfer is simultaneously induced through the application of off-resonance irradiation in between excitation and acquisition of an RF-spoiled gradient-echo scheme. Applying the binary spin-bath model, an analytical signal equation is derived and verified through Bloch simulations. Monte Carlo simulations were performed to analyze the method's performance. The estimation of the binary spin-bath parameters with B 1 + $$ {B}_1^{+} $$ compensation was further investigated through experiments, both ex vivo and in vivo. RESULTS: Comparing BTS with existing methods, simulations showed that existing methods can significantly bias T 1 $$ {T}_1 $$ estimation when not accounting for transmit B 1 $$ {B}_1 $$ heterogeneity and MT effects that are present. Phantom experiments further showed that the degree of this bias increases with increasing macromolecular proton fraction. Multi-parameter fit results from an in vivo brain study generated values in agreement with previous literature. Based on these studies, we confirmed that BTS is a robust method for estimating the binary spin-bath parameters in macromolecule-rich environments, even in the presence of B 1 + $$ {B}_1^{+} $$ inhomogeneity. CONCLUSION: A method of estimating Bloch-Siegert shift and magnetization transfer effect has been developed and validated. Both simulations and experiments confirmed that BTS can estimate spin-bath parameters ( T 1 F $$ {T}_1^{\mathrm{F}} $$ , f $$ f $$ , k F $$ {k}_{\mathrm{F}} $$ ) that are free from B 1 + $$ {B}_1^{+} $$ bias.

Data adaptive regularization with reference tissue constraints for liver quantitative susceptibility mapping.

PURPOSE: To improve repeatability and reproducibility across acquisition parameters and reduce bias in quantitative susceptibility mapping (QSM) of the liver, through development of an optimized regularized reconstruction algorithm for abdominal QSM. METHODS: An optimized approach to estimation of magnetic susceptibility distribution is formulated as a constrained reconstruction problem that incorporates estimates of the input data reliability and anatomical priors available from chemical shift-encoded imaging. The proposed data-adaptive method was evaluated with respect to bias, repeatability, and reproducibility in a patient population with a wide range of liver iron concentration (LIC). The proposed method was compared to the previously proposed and validated approach in liver QSM for two multi-echo spoiled gradient-recalled echo protocols with different acquisition parameters at 3T. Linear regression was used for evaluation of QSM methods against a reference FDA-approved R 2 $$ {R}_2 $$ -based LIC measure and R 2 ∗ $$ {R}_2^{\ast } $$ measurements; repeatability/reproducibility were assessed by Bland-Altman analysis. RESULTS: The data-adaptive method produced susceptibility maps with higher subjective quality due to reduced shading artifacts. For both acquisition protocols, higher linear correlation with both R 2 $$ {R}_2 $$ - and R 2 ∗ $$ {R}_2^{\ast } $$ -based measurements were observed for the data-adaptive method ( r 2 = 0 . 74 / 0 . 69 $$ {r}^2=0.74/0.69 $$ for R 2 $$ {R}_2 $$ , 0 . 97 / 0 . 95 $$ 0.97/0.95 $$ for R 2 ∗ $$ {R}_2^{\ast } $$ ) than the standard method ( r 2 = 0 . 60 / 0 . 66 $$ {r}^2=0.60/0.66 $$ and 0 . 79 / 0 . 88 $$ 0.79/0.88 $$ ). For both protocols, the data-adaptive method enabled better test-retest repeatability (repeatability coefficients 0.19/0.30 ppm for the data-adaptive method, 0.38/0.47 ppm for the standard method) and reproducibility across protocols (reproducibility coefficient 0.28 vs. 0.53ppm) than the standard method. CONCLUSIONS: The proposed data-adaptive QSM algorithm may enable quantification of LIC with improved repeatability/reproducibility across different acquisition parameters as 3T.

A proposed new taxonomy of cognitive phenotypes in multiple sclerosis: The International Classification of Cognitive Disorders in MS (IC-CoDiMS).

BACKGROUND: Characterization of cognitive impairment (CI) in multiple sclerosis into distinct phenotypes holds promise for individualized treatments and biomarker exploration. OBJECTIVE: Apply a previously validated, neuropsychologically driven diagnostic algorithm to identify a taxonomy of the type of cognitive phenotypes in multiple sclerosis. METHODS: An algorithm developed and validated in other neurological diseases was applied to a cohort of 1281 people with multiple sclerosis who underwent clinical neuropsychological evaluation across three multiple sclerosis centers. A domain was marked impaired if scores on two tests within the domain fell below one of the two thresholds of interest (compared to controls; -1.0 SD and -1.5 SD below the mean). Results were then tabulated for each participant to determine the type of impairments across the sample. RESULTS: At -1 SD threshold, 48.7% were intact, 21.6% had single-domain, 14.3% bi-domain, and 15.4% multi-domain impairment. At -1.5 SD threshold, 72.9% were intact, 14.0% had single-domain, 8.2% bi-domain, and 5.0% multi-domain impairment. Processing speed was the most frequent single-domain impairment, followed by executive function and memory. CONCLUSIONS: These findings advance the taxonomy of cognitive phenotypes in multiple sclerosis and clarify the type and distribution of possible cognitive diagnoses, pave the way for further investigation of associated biomarkers, and provide clinically meaningful information to guide individualized treatment and rehabilitation.

High-resolution dynamic susceptibility contrast perfusion imaging using higher-order temporal smoothness regularization.

PURPOSE: To improve image quality and resolution of dynamic susceptibility contrast perfusion weighted imaging (DSC-PWI) by developing acquisition and reconstruction methods exploiting the temporal regularity property of DSC-PWI signal. THEORY AND METHODS: A novel regularized reconstruction is proposed that recovers DSC-PWI series from interleaved segmented spiral k-space acquisition using higher order temporal smoothness (HOTS) properties of the DSC-PWI signal. The HOTS regularization is designed to tackle representational insufficiency of the standard first-order temporal regularizations for supporting higher accelerations. The higher accelerations allow for k-space coverage with shorter spiral interleaves resulting in improved acquisition point spread function, and acquisition of images at multiple TEs for more accurate DSC-PWI analysis. RESULTS: The methods were evaluated in simulated and in-vivo studies. HOTS regularization provided increasingly more accurate models for DSC-PWI than the standard first-order methods with either quadratic or robust norms at the expense of increased noise. HOTS DSC-PWI optimized for noise and accuracy demonstrated significant advantages over both spiral DSC-PWI without temporal regularization and traditional echo-planar DSC-PWI, improving resolution and mitigating image artifacts associated with long readout, including blurring and geometric distortions. In context of multi-echo DSC-PWI, the novel methods allowed ∼4.3× decrease of voxel volume, providing 2× number of TEs compared to the previously published results. CONCLUSIONS: Proposed HOTS reconstruction combined with dynamic spiral sampling represents a valid mechanism for improving image quality and resolution of DSC-PWI significantly beyond those available with established fast imaging techniques.

Wide Arterial Sparing Encephalo-Duro-Synangiosis for Moyamoya: Surgical Technique and Outcomes.

BACKGROUND: Moyamoya is managed by surgical revascularization, but no standardized method has yet been universally adopted. OBJECTIVE: To describe a new indirect bypass technique for pediatric moyamoya, wide arterial sparing encephalo-duro-synangiosis (WASEDS), which provides a much wider area of revascularization with minimal compromise to the middle meningeal arterial tree compared with traditional procedures. Initially used as a salvage technique after failed encephalo-duro-arterio-synangiosis, its success later motivated its use as a first-line procedure. METHODS: Clinical and radiographic records of patients who underwent WASEDS for moyamoya from 2009 to 2020 were reviewed. Brain perfusion relative cerebral blood volume on the side of the WASEDS procedure was calculated. Two-tailed paired t tests were performed to identify the statistically significant differences ( P ≤ .05). RESULTS: WASEDS was successfully performed on 8 patients for a total of 14 cerebral hemispheres. Age ranged from 2 to 25 years. There were no mortalities. The average clinical and radiographic follow-up was 49.79 months (range 2-126 months), demonstrating improvement in neurological condition and no postoperative stroke and significant diminution or cessation of transient ischemic attacks in all patients. Relative cerebral blood volume increased 9.24% after the WASEDS procedure ( P = .012). There were no neurological complications. There were 2 pseudomeningoceles related to the extensive dural openings. CONCLUSION: WASEDS is a safe and effective indirect revascularization technique for both primary and salvage techniques. It provides an extensive area of cortical revascularization with no compromise of the middle meningeal vasculature and subjective reports of early improvement in cognition and behavior. The main disadvantage is elevated risk of pseudomeningocele secondary to the large craniotomy.

Decreased basal ganglia and thalamic iron in early psychotic spectrum disorders are associated with increased psychotic and schizotypal symptoms.

Iron deficits have been reported as a risk factor for psychotic spectrum disorders (PSD). However, examinations of brain iron in PSD remain limited. The current study employed quantitative MRI to examine iron content in several iron-rich subcortical structures in 49 young adult individuals with PSD (15 schizophrenia, 17 schizoaffective disorder, and 17 bipolar disorder with psychotic features) compared with 35 age-matched healthy controls (HC). A parametric approach based on a two-pool magnetization transfer model was applied to estimate longitudinal relaxation rate (R1), which reflects both iron and myelin, and macromolecular proton fraction (MPF), which is specific to myelin. To describe iron content, a synthetic effective transverse relaxation rate (R2*) was modeled using a linear fitting of R1 and MPF. PSD patients compared to HC showed significantly reduced R1 and synthetic R2* across examined regions including the pallidum, ventral diencephalon, thalamus, and putamen areas. This finding was primarily driven by decreases in the subgroup with schizophrenia, followed by schizoaffective disorder. No significant group differences were noted for MPF between PSD and HC while for regional volume, significant reductions in patients were only observed in bilateral caudate, suggesting that R1 and synthetic R2* reductions in schizophrenia and schizoaffective patients likely reflect iron deficits that either occur independently or precede structural and myelin changes. Subcortical R1 and synthetic R2* were also found to be inversely related to positive symptoms within the PSD group and to schizotypal traits across the whole sample. These findings that decreased iron in subcortical regions are associated with PSD risk and symptomatology suggest that brain iron deficiencies may play a role in PSD pathology and warrant further study.

The Influence of Data-Driven Compressed Sensing Reconstruction on Quantitative Pharmacokinetic Analysis in Breast DCE MRI.

Radial acquisition with MOCCO reconstruction has been previously proposed for high spatial and temporal resolution breast DCE imaging. In this work, we characterize MOCCO across a wide range of temporal contrast enhancement in a digital reference object (DRO). Time-resolved radial data was simulated using a DRO with lesions in different PK parameters. The under sampled data were reconstructed at 5 s temporal resolution using the data-driven low-rank temporal model for MOCCO, compressed sensing with temporal total variation (CS-TV) and more conventional low-rank reconstruction (PCB). Our results demonstrated that MOCCO was able to recover curves with Ktrans values ranging from 0.01 to 0.8 min-1 and fixed Ve = 0.3, where the fitted results are within a 10% bias error range. MOCCO reconstruction showed less impact on the selection of different temporal models than conventional low-rank reconstruction and the greater error was observed with PCB. CS-TV showed overall underestimation in both Ktrans and Ve. For the Monte-Carlo simulations, MOCCO was found to provide the most accurate reconstruction results for curves with intermediate lesion kinetics in the presence of noise. Initial in vivo experiences are reported in one patient volunteer. Overall, MOCCO was able to provide reconstructed time-series data that resulted in a more accurate measurement of PK parameters than PCB and CS-TV.

Deep learning approach to predict pain progression in knee osteoarthritis.

OBJECTIVE: To develop and evaluate deep learning (DL) risk assessment models for predicting pain progression in subjects with or at risk of knee osteoarthritis (OA). MATERIALS AND METHODS: The incidence and progression cohorts of the Osteoarthritis Initiative, a multi-center longitudinal study involving 9348 knees in 4674 subjects with or at risk of knee OA that began in 2004 and is ongoing, were used to conduct this retrospective analysis. A subset of knees without and with pain progression (defined as a 9-point or greater increase in pain score between baseline and two or more follow-up time points over the first 48 months) was randomly stratified into training (4200 knees with a mean age of 61.0 years and 60% female) and hold-out testing (500 knees with a mean age of 60.8 years and 60% female) datasets. A DL model was developed to predict pain progression using baseline knee radiographs. An artificial neural network was used to develop a traditional risk assessment model to predict pain progression using demographic, clinical, and radiographic risk factors. A combined model was developed to combine demographic, clinical, and radiographic risk factors with DL analysis of baseline knee radiographs. Area under the curve (AUC) analysis was performed using the hold-out testing dataset to evaluate model performance. RESULTS: The traditional model had an AUC of 0.692 (66.9% sensitivity and 64.1% specificity). The DL model had an AUC of 0.770 (76.7% sensitivity and 70.5% specificity), which was significantly higher (p < 0.001) than the traditional model. The combined model had an AUC of 0.807 (72.3% sensitivity and 80.9% specificity), which was significantly higher (p < 0.05) than the traditional and DL models. CONCLUSIONS: DL models using baseline knee radiographs had higher diagnostic performance for predicting pain progression than traditional models using demographic, clinical, and radiographic risk factors.

Quantitative Macromolecular Proton Fraction Mapping Reveals Altered Cortical Myelin Profile in Schizophrenia Spectrum Disorders.

Myelin abnormalities have been reported in schizophrenia spectrum disorders (SSD) in white matter. However, in vivo examinations of cortical myeloarchitecture in SSD, especially those using quantitative measures, are limited. Here, we employed macromolecular proton fraction (MPF) obtained from quantitative magnetization transfer imaging to characterize intracortical myelin organization in 30 SSD patients versus 34 healthy control (HC) participants. We constructed cortical myelin profiles by extracting MPF values at various cortical depths and quantified their shape using a nonlinearity index (NLI). To delineate the association of illness duration with myelin changes, SSD patients were further divided into 3 duration groups. Between-group comparisons revealed reduced NLI in the SSD group with the longest illness duration (>5.5 years) compared with HC predominantly in bilateral prefrontal areas. Within the SSD group, cortical NLI decreased with disease duration and was positively associated with a measure of spatial working memory capacity as well as with cortical thickness (CT). Layer-specific analyses suggested that NLI decreases in the long-duration SSD group may arise in part from significantly increased MPF values in the midcortical layers. The current study reveals cortical myelin profile changes in SSD with illness progression, which may reflect an abnormal compensatory mechanism of the disorder.

Correlation of Objective Endpoints and Subjective Patient-Reported Outcomes in NAFLD Treatment with Essential Phospholipids: Real-World Data Based on Pooled Analysis of Observational Studies.

BACKGROUND: While no "gold-standard" pharmacotherapy for nonalcoholic fatty liver disease (NAFLD) is yet established, essential phospholipids (EPLs) are reported to decrease steatosis and improve laboratory parameters. OBJECTIVE: This analysis evaluated adherence and satisfaction with EPL treatment as patient-reported outcomes and their relationship with changes in laboratory and ultrasound parameters among Russian patients with NAFLD. METHODS: Data were pooled from three observational Russian studies-MANPOWER (2015-2016), LIDER 1 (2012-2013), and LIDER 2 (2013)-in which EPLs were used for at least 12 weeks in the treatment of liver diseases and which measured both subjective and objective endpoints. Only patients who had NAFLD were included in this analysis. The main endpoints were to determine treatment adherence and satisfaction with 12 weeks of EPL therapy, relationship between adherence/satisfaction and changes in the laboratory and ultrasound parameters. A secondary subgroup analysis was performed to identify patients with NAFLD who responded better (or worse) to 24 weeks of adjunctive EPL treatment. RESULTS: Overall, 3384 patients were included. A total of 82.2% of patients were adherent to 12 weeks of EPL treatment; high/very high satisfaction was reported by 15.3%/65.9% of clinicians and 15.9%/64.4% of patients. There was positive correlation between patients' adherence and satisfaction and significant improvement in laboratory (transaminases, lipid profile; p < 0.001) and ultrasound (steatosis, p < 0.001) parameters, and improvement in symptoms (p < 0.001) after 24 weeks of EPL. Male patients, patients with unhealthy lifestyles, and those with more comorbidities showed a better response in laboratory and ultrasound parameters. CONCLUSIONS: Patients with NAFLD treated with adjunctive EPL therapy in real-world clinical practice in Russia showed good treatment adherence and treatment satisfaction. Improvements in laboratory and ultrasound parameters, as well as dynamics of patient symptoms, were positively correlated with adherence and satisfaction.

Nonalcoholic fatty liver disease (NAFLD) is a major cause of liver disease, and patients have a risk of liver cancer and needing transplantation. Patients with advanced NAFLD are usually recommended to use medication. Clinical trials in patients with NAFLD showed that essential phospholipids (EPLs) comprising phosphatidylcholine improved liver fat accumulation (known as steatosis), so treatment guidelines in several countries recommend they be added to the usual therapy. EPLs have been associated with both objective and subjective improvements. This real-world study evaluated three observational studies to assess how patient adherence and satisfaction with EPL treatment were related to changes in clinical parameters in 3384 Russian patients with NAFLD. Overall, 82.2% of patients were adherent with 12 weeks of EPL therapy, and almost two-thirds of patients (64.4%) and clinicians (65.9%) reported very high satisfaction with treatment. Patients' adherence and satisfaction increased alongside significant improvements in liver enzymes, lipid levels, liver fat content (steatosis), and fewer symptoms after 24 weeks of EPL therapy. Male patients, those with an unhealthy lifestyle, and those with other comorbid conditions had the best response to EPL therapy. This study shows that Russian patients with NAFLD have good adherence and satisfaction with EPL therapy in routine clinical practice and highlights the importance of adherence to EPL therapy in these patients.

Bi-component T2 mapping correlates with articular cartilage material properties.

Non-invasive estimation of cartilage material properties is useful for understanding cartilage health and creating subject-specific computational models. Bi-component T2 mapping measured using Multi-Component Driven Equilibrium Single Shot Observation of T1 and T2 (mcDESPOT) is sensitive for detecting cartilage degeneration within the human knee joint, but has not been correlated with cartilage composition and mechanical properties. Therefore, the purpose of this study was to investigate the relationship between bi-component T2 parameters measured using mcDESPOT at 3.0 T and cartilage composition and mechanical properties. Ex-vivo patellar cartilage specimens harvested from five human cadaveric knees were imaged using mcDESPOT at 3.0 T. Cartilage samples were removed from the patellae, mechanically tested to determine linear modulus and dissipated energy, and chemically tested to determine proteoglycan and collagen content. Parameter maps of single-component T2 relaxation time (T2), the T2 relaxation times of the fast relaxing macromolecular bound water component (T2F) and slow relaxing bulk water component (T2S), and the fraction of the fast relaxing macromolecular bound water component (FF) were compared to mechanical and chemical measures using linear regression. FF was significantly (p < 0.05) correlated with energy dissipation and linear modulus. T2 was significantly (p ≤ 0.05) correlated with elastic modulus at 1 Hz and energy dissipated at all frequencies. There were no other significant (p = 0.13-0.97) correlations between mcDESPOT parameters and mechanical properties. FF was significantly (p = 0.04) correlated with proteoglycan content. There were no other significant (p = 0.19-0.92) correlations between mcDESPOT parameters and proteoglycan or collagen content. This study suggests that FF measured using mcDESPOT at 3.0 T could be used to non-invasively estimate cartilage proteoglycan content, elastic modulus, and energy dissipation.

Comparison of data-driven and general temporal constraints on compressed sensing for breast DCE MRI.

PURPOSE: Current breast DCE-MRI strategies provide high sensitivity for cancer detection but are known to be insufficient in fully capturing rapidly changing contrast kinetics at high spatial resolution across both breasts. Advanced acquisition and reconstruction strategies aim to improve spatial and temporal resolution and increase specificity for disease characterization. In this work, we evaluate the spatial and temporal fidelity of a modified data-driven low-rank-based model (known as MOCCO, model consistency condition) compressed-sensing (CS) reconstruction compared to CS with temporal total variation with radial acquisition for high spatial-temporal breast DCE MRI. METHODS: Reconstruction performance was characterized using numerical simulations of a golden-angle stack-of-stars breast DCE-MRI acquisition at 5-second temporal resolution. Specifically, MOCCO was compared with CS total variation and conventional SENSE reconstructions. The temporal model for MOCCO was prelearned over the source data, whereas CS total variation was performed using a first-order temporal gradient sparsity transform. RESULTS: The MOCCO reconstruction was able to capture rapid lesion kinetics while providing high image quality across a range of optimal regularization values. It also recovered kinetics in small lesions (1.5 mm) in line-profile analysis and error images, whereas g-factor maps showed relatively low and constant values with no significant artifacts. The CS-TV method demonstrated either recovery of high spatial resolution with reduced temporal accuracy using large regularization values, or recovery of rapid lesion kinetics with reduced image quality using low regularization values. CONCLUSION: Simulations demonstrated that MOCCO with radial acquisition provides a robust imaging technique for improving temporal fidelity, while maintaining high spatial resolution and image quality in the setting of bilateral breast DCE MRI.

Amyloid and Tau PET Imaging of Alzheimer Disease and Other Neurodegenerative Conditions.

Although diagnosing the syndrome of dementia is largely a clinical endeavor, neuroimaging plays an increasingly important role in accurately determining the underlying etiology, which extends beyond its traditional role in excluding other causes of altered cognition. New neuroimaging methods not only facilitate the diagnosis of the most common neurodegenerative conditions (particularly Alzheimer Disease [AD]) after symptom onset, but also show diagnostic promise even in the very early or presymptomatic phases of disease. Positron emission tomography (PET) is increasingly recognized as a key clinical tool for differentiating normal age-related changes in brain metabolism (using 18F-fluorodeoxyglucose [FDG]) from those seen in the earliest stages of specific forms of dementia. However, FDG PET only demonstrates nonspecific changes in altered parenchymal glucose uptake and not the specific etiologic proteinopathy causing the abnormal glucose uptake. A growing class of radiotracers targeting specific protein aggregates for amyloid-ß (Aß) and tau are changing the way AD is diagnosed, as these radiotracers directly label the underlying disease pathology. As these pathology-specific radiotracers are currently making their way to the clinic, it is important for the clinical neuroradiologist to understand the underlying patterns of Aß and tau deposition in the context of AD (across its clinical continuum) and in other causes of dementia, as well as understand the implications of current research.

A New, Non-Invasive Scale for Steatosis Developed Using Real-World Data From Russian Outpatients to Aid in the Diagnosis of Non-Alcoholic Fatty Liver Disease.

INTRODUCTION: The current non-invasive tools for the diagnosis of non-alcoholic fatty liver disease (NAFLD) have methodological limitations. We aimed to develop a non-invasive scale to assist in the diagnosis of NAFLD. To achieve our aim, we conducted a secondary analysis of data from a large observational study conducted in Russia. METHODS: This retrospective analysis assessed the frequency of NAFLD in the population of patients in the DIREG_L_06725 study, an epidemiological, observational, cross-sectional, multicenter study performed in 50,145 outpatients from 16 Russian cities. Among the cohort of patients diagnosed with NAFLD, we identified factors associated with the risk of NAFLD. To develop a non-invasive tool for diagnosing NAFLD, we also determined the frequency of steatohepatitis. RESULTS: Our analysis included 48,297 patients; NAFLD was present in 20,281 patients (42.0%). The majority (64.1%) were women (80.3% post-menopause), and 87% had a body mass index (BMI) > 27.0 kg/m2. We developed a fully non-invasive scale (St-index) that showed a specificity of 91.4% for ruling in steatosis, and a sensitivity of 93.8% for ruling out steatosis. Multivariate regression analyses conducted in the subgroups of patients aged ≥ 12 and < 18 years and those with BMI < 25.0 kg/m2 produced area under the receiver operating characteristic (AUROC) curve values of 0.8243 and 0.7054, respectively. The factors most strongly associated with the development of NAFLD were age > 35 years, presence of type 2 diabetes mellitus, and a waist circumference/height ratio > 0.54. CONCLUSION: Our non-invasive steatosis scale, St-index, can help physicians diagnose NAFLD in high-risk patients in the absence of ultrasound data.

Deep learning-based fully automatic segmentation of wrist cartilage in MR images.

The study objective was to investigate the performance of a dedicated convolutional neural network (CNN) optimized for wrist cartilage segmentation from 2D MR images. CNN utilized a planar architecture and patch-based (PB) training approach that ensured optimal performance in the presence of a limited amount of training data. The CNN was trained and validated in 20 multi-slice MRI datasets acquired with two different coils in 11 subjects (healthy volunteers and patients). The validation included a comparison with the alternative state-of-the-art CNN methods for the segmentation of joints from MR images and the ground-truth manual segmentation. When trained on the limited training data, the CNN outperformed significantly image-based and PB-U-Net networks. Our PB-CNN also demonstrated a good agreement with manual segmentation (Sørensen-Dice similarity coefficient [DSC] = 0.81) in the representative (central coronal) slices with a large amount of cartilage tissue. Reduced performance of the network for slices with a very limited amount of cartilage tissue suggests the need for fully 3D convolutional networks to provide uniform performance across the joint. The study also assessed inter- and intra-observer variability of the manual wrist cartilage segmentation (DSC = 0.78-0.88 and 0.9, respectively). The proposed deep learning-based segmentation of the wrist cartilage from MRI could facilitate research of novel imaging markers of wrist osteoarthritis to characterize its progression and response to therapy.

SANTIS: Sampling-Augmented Neural neTwork with Incoherent Structure for MR image reconstruction.

PURPOSE: To develop and evaluate a novel deep learning-based reconstruction framework called SANTIS (Sampling-Augmented Neural neTwork with Incoherent Structure) for efficient MR image reconstruction with improved robustness against sampling pattern discrepancy. METHODS: With a combination of data cycle-consistent adversarial network, end-to-end convolutional neural network mapping, and data fidelity enforcement for reconstructing undersampled MR data, SANTIS additionally utilizes a sampling-augmented training strategy by extensively varying undersampling patterns during training, so that the network is capable of learning various aliasing structures and thereby removing undersampling artifacts more effectively and robustly. The performance of SANTIS was demonstrated for accelerated knee imaging and liver imaging using a Cartesian trajectory and a golden-angle radial trajectory, respectively. Quantitative metrics were used to assess its performance against different references. The feasibility of SANTIS in reconstructing dynamic contrast-enhanced images was also demonstrated using transfer learning. RESULTS: Compared to conventional reconstruction that exploits image sparsity, SANTIS achieved consistently improved reconstruction performance (lower errors and greater image sharpness). Compared to standard learning-based methods without sampling augmentation (e.g., training with a fixed undersampling pattern), SANTIS provides comparable reconstruction performance, but significantly improved robustness, against sampling pattern discrepancy. SANTIS also achieved encouraging results for reconstructing liver images acquired at different contrast phases. CONCLUSION: By extensively varying undersampling patterns, the sampling-augmented training strategy in SANTIS can remove undersampling artifacts more robustly. The novel concept behind SANTIS can particularly be useful for improving the robustness of deep learning-based image reconstruction against discrepancy between training and inference, an important, but currently less explored, topic.

Resolving estimation uncertainties of chemical shift encoded fat-water imaging using magnetization transfer effect.

PURPOSE: B0 field inhomogeneity may cause significant errors in chemical shift encoding-based fat-water (F/W) separation. We describe a new approach to improve its robustness using novel B0 field map pre-estimation. METHODS: Our method exploits insensitivity of fat to magnetization transfer effect, which allows generating fat-insensitive B0 field priors with full or partial spatial support using a low-resolution magnetization transfer-weighted scan. The full prior can be employed by most F/W separation methods for initialization or data demodulation. We also propose a modified region-growing algorithm in which the partial prior is utilized for its initial seeding. RESULTS: The magnetization transfer-based B0 priors significantly reduced F/W errors of three representative F/W separation methods in all cases. In cases with moderate B0 inhomogeneity, the full prior allowed error-free separation even with basic, voxel-independent processing. When coupled with methods exploiting B0 field smoothness, it significantly improved separation accuracy even in the presence of strong inhomogeneities. Seeding the region-growing with the partial prior significantly improved performance of F/W separation, including cases with spatially disconnected tissues. CONCLUSION: Magnetization transfer-based B0 field pre-estimation provides valuable prior information for F/W separation, which may significantly improve its robustness at the expense of nominal (< 5%-10%) scan time increase.