Prognostic significance of the novel immunonutritional marker of cholesterol-to-lymphocyte ratio in patients with non-metastatic breast cancer.

BACKGROUND: Although there is a strong correlation between the novel cholesterol-to-lymphocyte ratio (CLR) and tumor survival, its prognostic significance in breast cancer (BC) is unknown. After analyzing the relationship between CLR and the overall survival (OS) of patients with BC, we created a predictive model. METHODS: Following retrospective enrollment, 1316 patients with BC were randomized into two cohorts: validation (n = 392) and training (n = 924). Distinct factors within the training dataset were identified for OS by univariate and multivariate Cox analyses; two-tailed P-value < 0.05 were considered to indicate statistical significance. On this premise, we developed novel signals for survival prediction and utilized the calibration curve, receiver operating characteristic curves, and concordance index (C-index) to validate their efficacy across both datasets. RESULTS: Patients with BC were categorized into two categories with differing prognoses based on the CLR score [hazard ratio = 0.492; 95% confidence interval (CI): 0.286-0.846, P = 0.009]. A prediction nomogram was created based on multivariate analysis, which showed that N stage, postoperative pathological categorization, and CLR score were all independently correlated with OS. In the training [C-index = 0.831 (95% CI: 0.788-0.874)] and validation [C-index = 0.775 (95% CI: 0.694-0.856)] cohorts, the nomogram demonstrated favorable performance in predicting OS. In both the training and validation cohorts, it outperformed the traditional staging system [C-index = 0.702 (95% CI: 0.623-0.782)] and [C-index = 0.709 (95% CI: 0.570-0.847)]. The accurate prediction by the signature was further demonstrated by the time-dependent receiver operating characteristic curves. CONCLUSIONS: The novel immunonutritional marker CLR could function as a simplified, cost-effective, easily accessible, non-invasive, and readily promotive prognostic indicator for patients with early-stage BC and demonstrates superior predictive power than the traditional staging system.

Learning CT-free attenuation-corrected total-body PET images through deep learning.

OBJECTIVES: Total-body PET/CT scanners with long axial fields of view have enabled unprecedented image quality and quantitative accuracy. However, the ionizing radiation from CT is a major issue in PET imaging, which is more evident with reduced radiopharmaceutical doses in total-body PET/CT. Therefore, we attempted to generate CT-free attenuation-corrected (CTF-AC) total-body PET images through deep learning. METHODS: Based on total-body PET data from 122 subjects (29 females and 93 males), a well-established cycle-consistent generative adversarial network (Cycle-GAN) was employed to generate CTF-AC total-body PET images directly while introducing site structures as prior information. Statistical analyses, including Pearson correlation coefficient (PCC) and t-tests, were utilized for the correlation measurements. RESULTS: The generated CTF-AC total-body PET images closely resembled real AC PET images, showing reduced noise and good contrast in different tissue structures. The obtained peak signal-to-noise ratio and structural similarity index measure values were 36.92 ± 5.49 dB (p < 0.01) and 0.980 ± 0.041 (p < 0.01), respectively. Furthermore, the standardized uptake value (SUV) distribution was consistent with that of real AC PET images. CONCLUSION: Our approach could directly generate CTF-AC total-body PET images, greatly reducing the radiation risk to patients from redundant anatomical examinations. Moreover, the model was validated based on a multidose-level NAC-AC PET dataset, demonstrating the potential of our method for low-dose PET attenuation correction. In future work, we will attempt to validate the proposed method with total-body PET/CT systems in more clinical practices. CLINICAL RELEVANCE STATEMENT: The ionizing radiation from CT is a major issue in PET imaging, which is more evident with reduced radiopharmaceutical doses in total-body PET/CT. Our CT-free PET attenuation correction method would be beneficial for a wide range of patient populations, especially for pediatric examinations and patients who need multiple scans or who require long-term follow-up. KEY POINTS: • CT is the main source of radiation in PET/CT imaging, especially for total-body PET/CT devices, and reduced radiopharmaceutical doses make the radiation burden from CT more obvious. • The CT-free PET attenuation correction method would be beneficial for patients who need multiple scans or long-term follow-up by reducing additional radiation from redundant anatomical examinations. • The proposed method could directly generate CT-free attenuation-corrected (CTF-AC) total-body PET images, which is beneficial for PET/MRI or PET-only devices lacking CT image poses.

Flexible machine-learning interatomic potential for simulating structural disordering behavior of Li7La3Zr2O12 solid electrolytes.

Batteries based on solid-state electrolytes, including Li7La3Zr2O12 (LLZO), promise improved safety and increased energy density; however, atomic disorder at grain boundaries and phase boundaries can severely deteriorate their performance. Machine-learning (ML) interatomic potentials offer a uniquely compelling solution for simulating chemical processes, rare events, and phase transitions associated with these complex interfaces by mixing high scalability with quantum-level accuracy, provided that they can be trained to properly address atomic disorder. To this end, we report the construction and validation of an ML potential that is specifically designed to simulate crystalline, disordered, and amorphous LLZO systems across a wide range of conditions. The ML model is based on a neural network algorithm and is trained using ab initio data. Performance tests prove that the developed ML potential can predict accurate structural and vibrational characteristics, elastic properties, and Li diffusivity of LLZO comparable to ab initio simulations. As a demonstration of its applicability to larger systems, we show that the potential can correctly capture grain boundary effects on diffusivity, as well as the thermal transition behavior of LLZO. These examples show that the ML potential enables simulations of transitions between well-defined and disordered structures with quantum-level accuracy at speeds thousands of times faster than ab initio methods.

Cooperative insertion of CO2 in diamine-appended metal-organic frameworks.

The process of carbon capture and sequestration has been proposed as a method of mitigating the build-up of greenhouse gases in the atmosphere. If implemented, the cost of electricity generated by a fossil fuel-burning power plant would rise substantially, owing to the expense of removing CO2 from the effluent stream. There is therefore an urgent need for more efficient gas separation technologies, such as those potentially offered by advanced solid adsorbents. Here we show that diamine-appended metal-organic frameworks can behave as 'phase-change' adsorbents, with unusual step-shaped CO2 adsorption isotherms that shift markedly with temperature. Results from spectroscopic, diffraction and computational studies show that the origin of the sharp adsorption step is an unprecedented cooperative process in which, above a metal-dependent threshold pressure, CO2 molecules insert into metal-amine bonds, inducing a reorganization of the amines into well-ordered chains of ammonium carbamate. As a consequence, large CO2 separation capacities can be achieved with small temperature swings, and regeneration energies appreciably lower than achievable with state-of-the-art aqueous amine solutions become feasible. The results provide a mechanistic framework for designing highly efficient adsorbents for removing CO2 from various gas mixtures, and yield insights into the conservation of Mg(2+) within the ribulose-1,5-bisphosphate carboxylase/oxygenase family of enzymes.

Hydroxyl-Assisted Phosphorene Stabilization with Robust Device Performances.

Phosphorene (few-layer black phosphorus) has been widely investigated for its unique optical and electronic properties. However, it is challenging to synthesize and process stable phosphorene as it degrades rapidly upon exposure to oxygen and moisture under ambient conditions, which has limited its use in practical applications. Herein, we propose an alkali-assisted stabilization process to produce high-quality phosphorene nanosheets. Our morphology measurements show that alkali-treated phosphorene remains stable for over 7 days in air. Electrical measurements on alkali-treated BP devices further proved its stable electrical property under ambient conditions. We further demonstrate superior light-assisted electrochemical water splitting performance using stable phosphorene. We attribute the stabilization effect to the chemical modification of the surface of phosphorene with P-OH bond formation. This study paves the avenue for the implementation of phosphorene devices in ambient conditions.

Nanostructured Metal Hydrides for Hydrogen Storage.

Knowledge and foundational understanding of phenomena associated with the behavior of materials at the nanoscale is one of the key scientific challenges toward a sustainable energy future. Size reduction from bulk to the nanoscale leads to a variety of exciting and anomalous phenomena due to enhanced surface-to-volume ratio, reduced transport length, and tunable nanointerfaces. Nanostructured metal hydrides are an important class of materials with significant potential for energy storage applications. Hydrogen storage in nanoscale metal hydrides has been recognized as a potentially transformative technology, and the field is now growing steadily due to the ability to tune the material properties more independently and drastically compared to those of their bulk counterparts. The numerous advantages of nanostructured metal hydrides compared to bulk include improved reversibility, altered heats of hydrogen absorption/desorption, nanointerfacial reaction pathways with faster rates, and new surface states capable of activating chemical bonds. This review aims to summarize the progress to date in the area of nanostructured metal hydrides and intends to understand and explain the underpinnings of the innovative concepts and strategies developed over the past decade to tune the thermodynamics and kinetics of hydrogen storage reactions. These recent achievements have the potential to propel further the prospects of tuning the hydride properties at nanoscale, with several promising directions and strategies that could lead to the next generation of solid-state materials for hydrogen storage applications.

Hybrid DFT investigation of the energetics of Mg ion diffusion in α-MoO3.

Rechargeable batteries that utilize divalent Mg ions as the charge carrier species can in principle achieve substantially greater volumetric energy densities than conventional Li-ion batteries. One significant impediment to the development of commercially viable Mg-ion batteries is the slow rate of Mg ion diffusion through otherwise promising cathode materials. Accurate prediction of the activation energies associated with this diffusion process using density functional theory (DFT) is especially challenging due to self-interaction errors intrinsic to DFT that lead to over-delocalization of the d-electrons. One effective but highly computationally demanding approach to reducing self-interaction errors is the use of hybrid functionals, which incorporate a fraction of exact Hartree-Fock exchange. In this work, we assess the effects of exact exchange on computed activation energies for ion diffusion in one potential cathode material, α-MoO3. In contrast to previous studies that primarily utilize non-hybrid functionals, we perform nudged elastic band calculations in which the nuclear coordinates are fully converged using both hybrid functionals and k-point sampling. It is found that while non-hybrid functionals indicate the existence of thermodynamically accessible channels for bulk Mg ion diffusion in all three dimensions, hybrid functionals predict that some of these channels are largely inaccessible under typical charge/discharge conditions. Furthermore, it is demonstrated that certain commonly used approximations for incorporating the effects of Hartree-Fock exchange are inadequate for this system, including DFT+U calculations and the use of single-point hybrid calculations using atomic positions obtained using non-hybrid functionals.

Atomically Thin Interfacial Suboxide Key to Hydrogen Storage Performance Enhancements of Magnesium Nanoparticles Encapsulated in Reduced Graphene Oxide.

As a model system for hydrogen storage, magnesium hydride exhibits high hydrogen storage density, yet its practical usage is hindered by necessarily high temperatures and slow kinetics for hydrogenation-dehydrogenation cycling. Decreasing particle size has been proposed to simultaneously improve the kinetics and decrease the sorption enthalpies. However, the associated increase in surface reactivity due to increased active surface area makes the material more susceptible to surface oxidation or other side reactions, which would hinder the overall hydrogenation-dehydrogenation process and diminish the capacity. Previous work has shown that the chemical stability of Mg nanoparticles can be greatly enhanced by using reduced graphene oxide as a protecting agent. Although no bulklike crystalline MgO layer has been clearly identified in this graphene-encapsulated/Mg nanocomposite, we propose that an atomically thin layer of honeycomb suboxide exists, based on first-principles interpretation of Mg K-edge X-ray absorption spectra. Density functional theory calculations reveal that in contrast to conventional expectations for thick oxides this interfacial oxidation layer permits H2 dissociation to the same degree as pristine Mg metal with the added benefit of enhancing the binding between reduced graphene oxide and the Mg nanoparticle, contributing to improved mechanical and chemical stability of the functioning nanocomposite.

Revealing electronic structure changes in Chevrel phase cathodes upon Mg insertion using X-ray absorption spectroscopy.

Following previous work predicting the electronic response of the Chevrel phase Mo6S8 upon Mg insertion (Thöle et al., Phys. Chem. Chem. Phys., 2015, 17, 22548), we provide the experimental proof, evident in X-ray absorption spectroscopy, to illustrate the charge compensation mechanism of the Chevrel phase compound during Mg insertion and de-insertion processes.

Re-examining the Chevrel phase Mo6S8 cathode for Mg intercalation from an electronic structure perspective.

We re-examine the electronic response of the Chevrel phase Mo6S8 upon Mg intercalation. The ground state Mo6S8 is metallic and exhibits strongly localized electronic screening of Mg(2+) ions. This localized screening cloud effectively shields the 2+ charge carried by Mg ions on the length scale of one unit cell and facilitates Mg ion diffusion.

Probing the mechanism of CO2 capture in diamine-appended metal-organic frameworks using measured and simulated X-ray spectroscopy.

Diamine-appended metal-organic frameworks display great promise for carbon capture applications, due to unusual step-shaped adsorption behavior that was recently attributed to a cooperative mechanism in which the adsorbed CO2 molecules insert into the metal-nitrogen bonds to form ordered ammonium carbamate chains [McDonald et al., Nature, 2015, 519, 303]. We present a detailed study of this mechanism by in situ X-ray absorption spectroscopy and density functional theory calculations. Distinct spectral changes at the N and O K-edges are apparent upon CO2 adsorption in both mmen-Mg2(dobpdc) and mmen-Mn2(dobpdc), and these are evaluated based upon computed spectra from three potential adsorption structures. The computations reveal that the observed spectral changes arise from specific electronic states that are signatures of a quasi-trigonal planar carbamate species that is hydrogen bonded to an ammonium cation. This eliminates two of the three structures studied, and confirms the insertion mechanism. We note the particular sensitivity of X-ray absorption spectra to the insertion step of this mechanism, underpinning the strength of the technique for examining subtle chemical changes upon gas adsorption.

The solvation structure of Mg ions in dichloro complex solutions from first-principles molecular dynamics and simulated X-ray absorption spectra.

The knowledge of Mg solvation structure in the electrolyte is requisite to understand the transport behavior of Mg ions and their dissolution/deposition mechanism at electrolyte/electrode interfaces. In the first established rechargeable Mg-ion battery system [D. Aurbach et al. Nature 2000, 407, 724], the electrolyte is of the dichloro complex (DCC) solution family, Mg(AlCl2BuEt)2/THF, resulting from the reaction of Bu2Mg and EtAlCl2 with a molar ratio of 1:2. There is disagreement in the literature regarding the exact solvation structure of Mg ions in such solutions, i.e., whether Mg(2+) is tetra- or hexacoordinated by a combination of Cl(-) and THF. In this work, theoretical insight into the solvation complexes present is provided based on first-principles molecular dynamics simulations (FPMD). Both Mg monomer and dimer structures are considered in both neutral and positively charged states. We found that, at room temperature, the Mg(2+) ion tends to be tetracoordinated in the THF solution phase instead of hexacoordinated, which is the predominant solid-phase coordination. Simulating the X-ray absorption spectra (XAS) at the Mg K-edge by sampling our FPMD trajectories, our predicted solvation structure can be readily compared with experimental measurements. It is found that when changing from tetra- to hexacoordination, the onset of X-ray absorption should exhibit at least a 1 eV blue shift. We propose that this energy shift can be used to monitor changes in the Mg solvation sphere as it migrates through the electrolyte to electrolyte/electrode interfaces and to elucidate the mechanism of Mg dissolution/deposition.

Fusion of shallow and deep features from 18F-FDG PET/CT for predicting EGFR-sensitizing mutations in non-small cell lung cancer.

Background: Non-small cell lung cancer (NSCLC) patients with epidermal growth factor receptor-sensitizing (EGFR-sensitizing) mutations exhibit a positive response to tyrosine kinase inhibitors (TKIs). Given the limitations of current clinical predictive methods, it is critical to explore radiomics-based approaches. In this study, we leveraged deep-learning technology with multimodal radiomics data to more accurately predict EGFR-sensitizing mutations. Methods: A total of 202 patients who underwent both flourine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) scans and EGFR sequencing prior to treatment were included in this study. Deep and shallow features were extracted by a residual neural network and the Python package PyRadiomics, respectively. We used least absolute shrinkage and selection operator (LASSO) regression to select predictive features and applied a support vector machine (SVM) to classify the EGFR-sensitive patients. Moreover, we compared predictive performance across different deep models and imaging modalities. Results: In the classification of EGFR-sensitive mutations, the areas under the curve (AUCs) of ResNet-based deep-shallow features and only shallow features from different multidata were as follows: RES_TRAD, PET/CT vs. CT-only vs. PET-only: 0.94 vs. 0.89 vs. 0.92; and ONLY_TRAD, PET/CT vs. CT-only vs. PET-only: 0.68 vs. 0.50 vs. 0.38. Additionally, the receiver operating characteristic (ROC) curves of the model using both deep and shallow features were significantly different from those of the model built using only shallow features (P<0.05). Conclusions: Our findings suggest that deep features significantly enhance the detection of EGFR-sensitizing mutations, especially those extracted with ResNet. Moreover, PET/CT images are more effective than CT-only and PET-only images in producing EGFR-sensitizing mutation-related signatures.

Coupled Experimental-Theoretical Characterization of a Carbon Electrode in Vanadium Redox Flow Batteries using X-ray Absorption Spectroscopy.

Vanadium redox flow batteries (VRFBs) have emerged as promising solutions for stationary grid energy storage due to their high efficiency, scalability, safety, near room-temperature operation conditions, and the ability to independently size power and energy capacities. The performance of VRFBs heavily relies on the redox couple reactions of V2+/V3+ and VO2+/VO2+ on carbon electrodes. Therefore, a thorough understanding of the surface functionality of carbon electrodes and their propensity for degradation during electrochemical cycles is crucial for designing VRFBs with extended lifespans. In this study, we present a coupled experimental-theoretical approach based on carbon K edge X-ray absorption spectroscopy (XAS) to characterize carbon electrodes prepared under different conditions and identify relevant functional groups that contribute to unique spectroscopic features. Atomic models were created to represent functional groups, such as hydroxyl, carboxyl, methyl, and aldehyde, bonded to carbon atoms in either sp2 or sp3 environments. The interactions between functionalized carbon and various solvated vanadium complexes were modeled using density functional theory. A library of carbon K-edge XAS spectra was generated for distinct carbon atoms in different functional groups, both before and after interacting with solvated vanadium complexes. We demonstrate how these simulated spectra can be used to deconvolve ex situ experimental spectra measured from carbon electrodes and to track changes in the electrode composition following immersion in different electrolytes or extended cycling within a functional VRFB. By doing so, we identify the active species present on the carbon electrodes, which play a crucial role in determining their electrochemical performance.

A Novel Three-Channel Endorectal Coil for Prostate Magnetic Resonance Imaging at 3T.

OBJECTIVE: The purpose of this work is to develop a 3-channel endorectal coil (ERC-3C) structure to obtain higher signal-to-noise (SNR) and better parallel imaging performance for prostate magnetic resonance imaging (MRI) at 3T. METHODS: The coil performance was validated by in vivo studies and the SNR, g-factor, and diffusion-weighted imaging (DWI) were compared. A 2-channel endorectal coil (ERC-2C) with two orthogonal loops and a 12-channel external surface coil were employed for comparison. RESULTS: Compared with the ERC-2C with a quadrature configuration and the external 12-channel coil array, the proposed ERC-3C improved SNR performance by 23.9% and 428.9%, respectively. The improved SNR enables the ERC-3C to produce spatial high-resolution images of 0.24 mm × 0.24 mm × 2 mm (0.1152 µL) in the prostate area within 9 minutes. CONCLUSION: We developed an ERC-3C and validated its performance through in vivo MR imaging experiments. SIGNIFICANCE: The results demonstrated the feasibility of an ERC with more than two channels and that a higher SNR can be achieved using the ERC-3C compared with an orthogonal ERC-2C of the same coverage.

Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images.

BACKGROUND: Dynamic positron emission tomography (PET) images are useful in clinical practice because they can be used to calculate the metabolic parameters (Ki) of tissues using graphical methods (such as Patlak plots). Ki is more stable than the standard uptake value and has a good reference value for clinical diagnosis. However, the long scanning time required for obtaining dynamic PET images, usually an hour, makes this method less useful in some ways. There is a tradeoff between the scan durations and the signal-to-noise ratios (SNRs) of Ki images. The purpose of our study is to obtain approximately the same image as that produced by scanning for one hour in just half an hour, improving the SNRs of images obtained by scanning for 30 min and reducing the necessary 1-h scanning time for acquiring dynamic PET images. METHODS: In this paper, we use U-Net as a feature extractor to obtain feature vectors with a priori knowledge about the image structure of interest and then utilize a parameter generator to obtain five parameters for a two-tissue, three-compartment model and generate a time activity curve (TAC), which will become close to the original 1-h TAC through training. The above-generated dynamic PET image finally obtains the Ki parameter image. RESULTS: A quantitative analysis showed that the network-generated Ki parameter maps improved the structural similarity index measure and peak SNR by averages of 2.27% and 7.04%, respectively, and decreased the root mean square error (RMSE) by 16.3% compared to those generated with a scan time of 30 min. CONCLUSIONS: The proposed method is feasible, and satisfactory PET quantification accuracy can be achieved using the proposed deep learning method. Further clinical validation is needed before implementing this approach in routine clinical applications.

Feasibility of a deep learning algorithm to achieve the low-dose 68Ga-FAPI/the fast-scan PET images: a multicenter study.

OBJECTIVES: Our work aims to study the feasibility of a deep learning algorithm to reduce the 68Ga-FAPI radiotracer injected activity and/or shorten the scanning time and to investigate its effects on image quality and lesion detection ability. METHODS: The data of 130 patients who underwent 68Ga-FAPI positron emission tomography (PET)/CT in two centers were studied. Predicted full-dose images (DL-22%, DL-28% and DL-33%) were obtained from three groups of low-dose images using a deep learning method and compared with the standard-dose images (raw data). Injection activity for full-dose images was 2.16 ± 0.61 MBq/kg. The quality of the predicted full-dose PET images was subjectively evaluated by two nuclear physicians using a 5-point Likert scale, and objectively evaluated by the peak signal-to-noise ratio, structural similarity index and root mean square error. The maximum standardized uptake value and the mean standardized uptake value (SUVmean) were used to quantitatively analyze the four volumes of interest (the brain, liver, left lung and right lung) and all lesions, and the lesion detection rate was calculated. RESULTS: Data showed that the DL-33% images of the two test data sets met the clinical diagnosis requirements, and the overall lesion detection rate of the two centers reached 95.9%. CONCLUSION: Through deep learning, we demonstrated that reducing the 68Ga-FAPI injected activity and/or shortening the scanning time in PET/CT imaging was feasible. In addition, 68Ga-FAPI dose as low as 33% of the standard dose maintained acceptable image quality. ADVANCES IN KNOWLEDGE: This is the first study of low-dose 68Ga-FAPI PET images from two centers using a deep learning algorithm.

First-Principles Elucidation of Initial Dehydrogenation Pathways in Mg(BH4)2.

Complex borohydrides such as Mg(BH4)2 offer one of highest capacities to chemically store hydrogen for onboard applications; however, it suffers greatly from kinetic constraints that prevent realization of full capacity and reversibility. Understanding these kinetic limitations solely from experiments is extremely challenging due to the unusual complexity of various competing elemental reaction steps involved during the de/rehydrogenation reaction. This work aims to map out the energetics associated with initial dehydrogenation of Mg(BH4)2 from first-principles simulations and to identify the preferred reaction pathways. Our calculations suggest the rate-limiting step during BH4--B3H8- conversion is the formation of the B2H7- intermediate. We further emphasize and clarify that the B3H8- and H- intermediates, formed during initial Mg(BH4)2 decomposition, appear as molecular species that are embedded in the Mg-BH4-Mg matrix as evidenced in the nuclear magnetic resonance measurements and not as bulk MgH2 and Mg(B3H8)2 as previously assumed in theoretical predictions of the thermodynamics.

Automated morphologic analysis of intracranial and extracranial arteries using convolutional neural networks.

OBJECTIVE: To develop an automated method for 3D magnetic resonance (MR) vessel wall image analysis to facilitate morphologic quantification of intra- and extracranial arteries, including vessel centerline tracking, vessel straightening and reformation, vessel wall segmentation based on convoluted neural networks (CNNs), and morphological measurement. METHODS: MR vessel wall images acquired using DANTE-SPACE sequences and corresponding time-of-flight-MRA images of 67 subjects (including 47 healthy volunteers and 20 patients with atherosclerosis) were included in this study. The centerline of the vessel was firstly extracted from the MRA images and copyed to the 3D MR vessel wall images through the registration relationship between the MRA images and the MR vessel wall images to extract, straighten, and reconstruct interested vessel segments into 2D slices. Then a complete CNN-based Unet-like method was used to automatically segment the vessel wall to obtain quantitative morphological measurements such as maximum wall thicknesses and normalized wall index (NWI). RESULTS: The proposed automatic segmentation network was trained and validated with 11,735 slices and tested on 2517 slices. The method showed satisfactory agreement with manual segmentation method. The Dice coefficients of intracranial arteries were 0.90 for lumen and 0.78 for vessel wall, while the Dice coefficients of extracranial arteries were 0.90 for lumen and 0.82 for vessel wall. The maximum wall thickness and NWI obtained from the proposed automatic method were slightly larger than those obtained from the manual method for both intra- and extracranial arteries. However, there was no significant difference of the quantitative measurements between the two methods (p > 0.05). In addition, the automatically measured NWI of plaque slice was significantly larger than that of normal slice. CONCLUSION: We propose an automatic analysis method of MR vessel wall images, which can realize automatic segmentation of intra- and extracranial vessel wall. It is expected to facilitate large-scale arterial vessel wall morphological quantification. ADVANCES IN KNOWLEDGE: We have proposed an automatic method for analysis of intra- and extracranial MR vessel wall images simultaneously based on CNN, which can facilitate large-scale quantitative analyses of vessel walls.

Deep Learning-Based Automated Detection of Arterial Vessel Wall and Plaque on Magnetic Resonance Vessel Wall Images.

Purpose: To develop and evaluate an automatic segmentation method of arterial vessel walls and plaques, which is beneficial for facilitating the arterial morphological quantification in magnetic resonance vessel wall imaging (MRVWI). Methods: MRVWI images acquired from 124 patients with atherosclerotic plaques were included. A convolutional neural network-based deep learning model, namely VWISegNet, was used to extract the features from MRVWI images and calculate the category of each pixel to facilitate the segmentation of vessel wall. Two-dimensional (2D) cross-sectional slices reconstructed from all plaques and 7 main arterial segments of 115 patients were used to build and optimize the deep learning model. The model performance was evaluated on the remaining nine-patient test set using the Dice similarity coefficient (DSC) and average surface distance (ASD). Results: The proposed automatic segmentation method demonstrated satisfactory agreement with the manual method, with DSCs of 93.8% for lumen contours and 86.0% for outer wall contours, which were higher than those obtained from the traditional U-Net, Attention U-Net, and Inception U-Net on the same nine-subject test set. And all the ASD values were less than 0.198 mm. The Bland-Altman plots and scatter plots also showed that there was a good agreement between the methods. All intraclass correlation coefficient values between the automatic method and manual method were greater than 0.780, and greater than that between two manual reads. Conclusion: The proposed deep learning-based automatic segmentation method achieved good consistency with the manual methods in the segmentation of arterial vessel wall and plaque and is even more accurate than manual results, hence improved the convenience of arterial morphological quantification.