Intrinsic Stress Field for Liquid Surfaces.

The microscopic stress field inhomogeneity in the interfacial region adjacent to the liquid surface is the fundamental origin of the liquid surface tension, but because of broadening due to capillary fluctuations, a detailed molecular level understanding of the stress field remains elusive. In this work, we deconvolute the capillary fluctuations to reveal the intrinsic stress field and show that the atomic-level contributions to the surface tension are similar in functional form across a variety of monatomic systems. These contributions are confined to an interfacial region approximately 1.5±0.1 times the particle diameter for all systems studied. In addition, the intrinsic density and stress profiles show a strong spatial correlation that should be useful in the development of a statistical mechanical theory for the prediction of surface stress and surface tension.

Associations of sedentary behaviors with mental health outcomes in a cohort of patients with minor ischemic stroke.

OBJECTIVE: The relationship between sedentary behaviors and functional outcomes of acute ischemic stroke (AIS) has been previously reported. However, it remains unclear whether sedentary behaviors are associated with mental health outcomes in AIS patients. Therefore, the objective of this study was to investigate the mental health outcomes in patients with minor AIS one year after stroke onset. METHODS: This cross-sectional study recruited 1230 patients with minor AIS (NIHSS ≤ 5) from three hospitals in China. One year after discharge, patients were interviewed using face-to-face questionnaires, including the PHQ-9, GAD-7, and ISI, to assess symptoms of depression, anxiety, and insomnia, respectively. Participants were categorized into the long sedentary time group and the short sedentary time group based on the median sedentary time of all participants. The associations between leisure sedentary time and mental health outcomes were investigated. RESULTS: Participants with a long leisure sedentary time had higher PHQ-9, GAD-7, and ISI scores than those with a short sedentary time. Longer sedentary time was associated with an increased risk of experiencing symptoms of major depression (RR, 95% CI: 1.79, 1.47 to 2.18), anxiety (RR, 95% CI: 3.28, 2.08 to 5.18), and insomnia (RR, 95% CI: 2.58, 2.03 to 3.28) one year after a minor AIS. CONCLUSION: Excessive sedentary time is associated with long-term mental health conditions after stroke. Therefore, reducing the sedentary time might be helpful for preventing poststroke depression, anxiety, and insomnia.

A case-control study of the risk factors for fistula-in-ano in infants and toddlers.

BACKGROUND: Limited data are available regarding the risk factors for fistula-in-ano (FIA) in infants and toddlers, potentially affecting their daily lives. OBJECTIVES: The purpose of this study was to identify potential risk factors for FIA in infants and toddlers, in order to implement early preventive interventions, avoid disease progression, and develop therapeutic strategies. DESIGN AND SETTINGS: A retrospective case-control study was conducted, comparing 41 infants and toddlers diagnosed with FIA with 41 healthy controls, between August 2020 and December 2021. INDEPENDENT VARIABLES: (a) maternal characteristics during pregnancy and delivery, (b) perinatal characteristics, dietary behaviors, and defecation-related behaviors in infants and toddlers, (c) family dietary behaviors. RESULTS: Mothers of infants and toddlers with FIA had given birth more times in the past, while the infants and toddlers themselves had less mealtime, a higher rate of exclusive breastfeeding, frequent loose stools, and a larger proportion of used wipes, experiencing perianal skin anomalies. The logistic regression analysis revealed that there are four significant risk factors associated with the development of FIA in infants and toddlers, including the number of previous deliveries by the mother (OR 6.327), defecation frequency score (OR 5.351), stool consistency score (OR 5.017), and cleaning with wipes after defecation (OR 8.089). CONCLUSION: Based on our data, it appeared that FIA in infants and toddlers could be attributed to several factors. These included an increased number of previous deliveries by mothers, frequent loose stools, and repeated wipe use. To prevent the occurrence and worsening of the disease, it is important to improve the frequency and consistency of stooling and provide proper care. Further research is required to verify these findings in other clinical settings.

Engulfment and Pushing of Cylindrical Liquid Nano-Inclusion by Advancing Crystal/Melt Interface: An Atomistic Simulation Study.

We reported a molecular dynamics (MD) simulation study of an advancing pure Al(100)/melt interface that encounters a foreign immiscible liquid Pb cylindrical nano-inclusion. When the advancing interface approaches the inclusion, the interface may engulf, push to an extent and then engulf or push the nano-inclusion away from the solidifying phase depending on the velocity of the interface. Here, we investigated cylindrical liquid Pb nano-inclusion pushing or engulfment by a growing crystal Al that strongly depends on the velocity of the crystal/melt interface, and a critical velocity (vc) is deduced. If the velocity of the interface is less than vc, then the inclusion is pushed and engulfed otherwise. The relationship between vc and the radius of the nano-inclusion is expressed using a power function that agrees well with the previous studies. For velocity above the vc, the crystal/melt interface plays a vital role; it hinders the matrix atoms from setting below the cylindrical nano-inclusion due to insufficient mass transfer below the inclusion, resulting in the engulfment.

Rare rectal gastrointestinal stromal tumor case: A case report and review of the literature.

BACKGROUND: Gastrointestinal stromal tumors (GISTs) are rare tumors of the gastrointestinal tract accounting for less than 1% of all gut tumors. GISTs occurring in the rectum are extremely rare, and these usually present at an advanced stage compared with other sites. CASE SUMMARY: A 60-year-old male who presented with features of sensations of rectal tenesmus was referred to our department with a mass in the lower rectum that was detected during a routine checkup. Colonoscopy, transrectal ultrasound, perianal magnetic resonance imaging and ultrasonic contrast were used to diagnose the rectum GIST, and then the patient underwent complete transanal resection using the ultrasonic scalpel. The patient was discharged ten days after the operation and was defined as low risk. Therefore, he had no need to receive subsequent adjuvant therapies, and he had not suffered any anal dysfunction or had any evidence of recurrence at follow up. CONCLUSION: Surgical resection with histologically negative margins is the standard curative treatment for rectal GISTs. Appropriate surgical techniques based on the location, size, and resectability of the tumor should attract great attention from clinicians.

Functional brain network identification and fMRI augmentation using a VAE-GAN framework.

Recently, deep learning models have achieved superior performance for mapping functional brain networks from functional magnetic resonance imaging (fMRI) data compared with traditional methods. However, due to the lack of sufficient data and the high dimensionality of brain volume, deep learning models of fMRI tend to suffer from overfitting. In addition, existing methods rarely studied fMRI data augmentation and its application. To address these issues, we developed a VAE-GAN framework that combined a VAE (variational auto-encoder) with a GAN (generative adversarial net) for functional brain network identification and fMRI augmentation. As a generative model, the VAE-GAN models the distribution of fMRI so that it enables the extraction of more generalized features, and thus relieve the overfitting issue. The VAE-GAN is easier to train on fMRI than a standard GAN since it uses latent variables from VAE to generate fake data rather than relying on random noise that is used in a GAN, and it can generate higher quality of fake data than VAE since the discriminator can promote the training of the generator. In other words, the VAE-GAN inherits the advantages of VAE and GAN and avoids their limitations in modeling of fMRI data. Extensive experiments on task fMRI datasets from HCP have proved the effectiveness and superiority of the proposed VAE-GAN framework for identifying both temporal features and functional brain networks compared with existing models, and the quality of fake data is higher than those from VAE and GAN. The results on resting state fMRI of Attention Deficit Hyperactivity Disorder (ADHD)-200 dataset further demonstrated that the fake data generated by the VAE-GAN can help improve the performance of brain network modeling and ADHD classification.

Excessive Sedentary Time Is Associated with Cognitive Decline in Older Patients with Minor Ischemic Stroke.

BACKGROUND: Cognitive impairment is commonly seen after acute ischemic stroke (AIS). Sedentary behaviors increase the risk of dementia among community dwelling population. OBJECTIVE: This study aims to investigate the association of sedentary behaviors with poststroke cognitive impairment among older adults with minor AIS. METHODS: This cohort study recruited 594 older subjects with minor AIS from three hospitals in China during February 1, 2016, and December 31, 2018. Participants were followed up for two years and the sedentary time per day was self-reported at the end of follow-up. Cognitive functions were assessed by Mini-Mental State Examination (MMSE). Participants were categorized into the high and low sedentary time group according to the median sedentary time of the participants. RESULTS: At two years of follow-up, the long sedentary time group had significantly lower MMSE scores than the short sedentary time group [median, (IQR): 21 (18 to 25) versus 22 (18 to 25), p = 0.368]. The long sedentary time group had a higher speed of cognitive decline than the short sedentary time group. Excessive sedentary time was associated with a higher risk of longitudinal cognitive decline (OR: 2.267, 95% CI: 1.594 to 3.225), adjusting for age, sex, education, body mass index, APOE genotype, comorbidities, symptoms of depression, anxiety, and insomnia, baseline MMSE scores and National Institute of Health Stroke Scale scores, cognitive therapy, and TOAST ischemic stroke subtypes. CONCLUSIONS: This study identified a possible link between sedentary behaviors and longitudinal cognitive decline among older patients with minor AIS, suggesting that reducing sedentary time might be helpful for preventing poststroke dementia.

Research on Wind Turbine Fault Detection Based on the Fusion of ASL-CatBoost and TtRSA.

The internal structure of wind turbines is intricate and precise, although the challenging working conditions often give rise to various operational faults. This study aims to address the limitations of traditional machine learning algorithms in wind turbine fault detection and the imbalance of positive and negative samples in the fault detection dataset. To achieve the real-time detection of wind turbine group faults and to capture wind turbine fault state information, an enhanced ASL-CatBoost algorithm is proposed. Additionally, a crawling animal search algorithm that incorporates the Tent chaotic mapping and t-distribution mutation strategy is introduced to assess the sensitivity of the ASL-CatBoost algorithm toward hyperparameters and the difficulty of manual hyperparameter setting. The effectiveness of the proposed hyperparameter optimization strategy, termed the TtRSA algorithm, is demonstrated through a comparison of traditional intelligent optimization algorithms using 11 benchmark test functions. When applied to the hyperparameter optimization of the ASL-CatBoost algorithm, the TtRSA-ASL-CatBoost algorithm exhibits notable enhancements in accuracy, recall, and other performance measures compared with the ASL-CatBoost algorithm and other ensemble learning algorithms. The experimental results affirm that the proposed algorithm model improvement strategy effectively enhances the wind turbine fault detection classification recognition rate.

A deep learning method for autism spectrum disorder identification based on interactions of hierarchical brain networks.

BACKGROUND: It has been recently shown that deep learning models exhibited remarkable performance of representing functional Magnetic Resonance Imaging (fMRI) data for the understanding of brain functional activities. With hierarchical structure, deep learning models can infer hierarchical functional brain networks (FBN) from fMRI. However, the applications of the hierarchical FBNs have been rarely studied. METHODS: In this work, we proposed a hierarchical recurrent variational auto-encoder (HRVAE) to unsupervisedly model the fMRI data. The trained HRVAE encoder can predict hierarchical temporal features from its three hidden layers, and thus can be regarded as a hierarchical feature extractor. Then LASSO (least absolute shrinkage and selection operator) regression was applied to estimate the corresponding hierarchical FBNs. Based on the hierarchical FBNs from each subject, we constructed a novel classification framework for brain disorder identification and test it on the Autism Brain Imaging Data Exchange (ABIDE) dataset, a world-wide multi-site database of autism spectrum disorder (ASD). We analyzed the hierarchy organization of FBNs, and finally used the overlaps of hierarchical FBNs as features to differentiate ASD from typically developing controls (TDC). RESULTS: The experimental results on 871 subjects from ABIDE dataset showed that the HRVAE model can effectively derive hierarchical FBNs including many well-known resting state networks (RSN). Moreover, the classification result improved the state-of-the-art by achieving a very high accuracy of 82.1 %. CONCLUSIONS: This work presents a novel data-driven deep learning method using fMRI data for ASD identification, which could provide valuable reference for clinical diagnosis. The classification results suggest that the interactions of hierarchical FBNs have association with brain disorder, which promotes the understanding of FBN hierarchy and could be applied to other brain disorder analysis.

Dynamic surface stress field of the pure liquid-vapor interface subjected to the cyclic loads.

We demonstrate a methodology for computationally investigating the mechanical response of a pure molten lead surface system to the lateral mechanical cyclic loads and try to answer the following question: how does the dynamically driven liquid surface system follow the classical physics of the elastic-driven oscillation? The steady-state oscillation of the dynamic surface tension (or excess stress) under cyclic load, including the excitation of high-frequency vibration mode at different driving frequencies and amplitudes, was compared with the classical theory of a single-body driven damped oscillator. Under the highest studied frequency (50 GHz) and amplitude (5%) of the load, the increase of in (mean value) dynamic surface tension could reach ∼5%. The peak and trough values of the instantaneous dynamic surface tension could reach (up to) 40% increase and (up to) 20% decrease compared to the equilibrium surface tension, respectively. The extracted generalized natural frequencies seem to be intimately related to the intrinsic timescales of the atomic temporal-spatial correlation functions of the liquids both in the bulk region and in the outermost surface layers. These insights uncovered could be helpful for quantitative manipulation of the liquid surface using ultrafast shockwaves or laser pulses.

A Systematic Review and Meta-Analysis of Comparing Drainage Alone versus Drainage with Primary Fistula Treatment for the Perianal Abscess in Children.

This systematic review and meta-analysis of nonrandomized studies (NRSs) aimed to evaluate the clinical efficacy and safety of two types of surgical interventions (respectively drainage alone and drainage with primary fistula treatment) for perianal abscesses (PAs) in children. Studies from 1992 to July 2022 were searched in 10 electronic databases. All relevant NRSs with available data which compared surgical drainage with or without primary fistula treatment were included. Patients with underlying diseases which led to abscess formation were excluded. The Newcastle-Ottawa Scale was used to assess the risk of bias and quality of the included studies. The outcomes were the healing rate, fistula formation rate, fecal incontinence, and wound healing duration. A total of 16 articles with 1,262 patients were considered suitable for the final meta-analysis. Primary fistula treatment was associated with a significantly higher healing rate when compared with incision and drainage alone (odds ratio [OR]: 5.76, 95% confidence interval [CI]: 4.04-8.22). This aggressive procedure for PA resulted in an 86% reduction in the fistula formation rate (OR: 0.14, 95% CI: 0.06-0.32). Limited data showed patients who underwent primary fistula treatment have a minor effect on postoperative fecal incontinence. Primary fistula treatment demonstrates a better clinical efficacy in promoting the healing rate and decreasing the formation of fistulas in PAs in children. The available evidence for a minor impact on anal function after this intervention is less strong.

Can transcutaneous perianal ultrasonography be the first-line diagnostic instrument for evaluating pediatric perianal fistulas?

Background: Pediatric perianal fistula is a common disorder. It is more difficult to detect the fistula tract and internal opening (IO) in children than in adults. This study aimed to evaluate the clinical diagnostic value of transcutaneous perianal ultrasound for children with perianal fistula. Methods: A retrospective review was conducted by analysing the preoperative transcutaneous perianal ultrasound and intraoperative exploration results of 203 consecutive patients who were <3 years old and diagnosed with perianal fistula. Analyses were conducted to evaluate the accuracy and consistency of utilizing the transcutaneous perianal ultrasound in the diagnosis of the complexity and location of the IO of perianal fistulas. Results: Compared with intraoperative exploration, the preoperative transcutaneous perianal ultrasonography has almost perfect agreement (Kappa = 0.881, P < 0.001) in the diagnosis of fistula tract complexity and IO with a sensitivity of 92% and a specificity of 97%. In addition, both intraoperative exploration and transcutaneous perianal ultrasound diagnosis showed high consistency in the identification of the IO of perianal fistulas (Quadrant I Kappa = 0.831, Quadrant II Kappa = 0.773, Quadrant III Kappa = 0.735, Quadrant IV Kappa = 0.802, all P < 0.01). The IOs were mainly distributed in Quadrants IV and II in both simple and complex fistulas. Conclusions: Transcutaneous perianal ultrasound, as a non-invasive and simple imaging technique, showed high accuracy in the diagnosis and identification of the fistula classification and IO location. It could be considered a first-line diagnostic instrument for evaluating perianal fistulas among children.

Atomistic characterization of the SiO2 high-density liquid/low-density liquid interface.

The equilibrium silica liquid-liquid interface between the high-density liquid (HDL) phase and the low-density liquid (LDL) phase is examined using molecular-dynamics simulation. The structure, thermodynamics, and dynamics within the interfacial region are characterized in detail and compared with previous studies on the liquid-liquid phase transition (LLPT) in bulk silica, as well as traditional crystal-melt interfaces. We find that the silica HDL-LDL interface exhibits a spatial fragile-to-strong transition across the interface. Calculations of dynamics properties reveal three types of dynamical heterogeneity hybridizing within the silica HDL-LDL interface. We also observe that as the interface is traversed from HDL to LDL, the Si/O coordination number ratio jumps to an unexpectedly large value, defining a thin region of the interface where HDL and LDL exhibit significant mixing. In addition, the LLPT phase coexistence is interpreted in the framework of the traditional thermodynamics of alloys and phase equilibria.

Local collective dynamics at equilibrium BCC crystal-melt interfaces.

We present a classical molecular-dynamics study of the collective dynamical properties of the coexisting liquid phase at equilibrium body-centered cubic (BCC) Fe crystal-melt interfaces. For the three interfacial orientations (100), (110), and (111), the collective dynamics are characterized through the calculation of the intermediate scattering functions, dynamical structure factors, and density relaxation times in a sequential local region of interest. An anisotropic speedup of the collective dynamics in all three BCC crystal-melt interfacial orientations is observed. This trend differs significantly from the previously observed slowing down of the local collective dynamics at the liquid-vapor interface [del Rio and González, Acta Mater. 198, 281 (2020)]. Examining the interfacial density relaxation times, we revisit the validity of the recently developed time-dependent Ginzburg-Landau theory for the solidification crystal-melt interface kinetic coefficients, resulting in excellent agreement with both the magnitude and the kinetic anisotropy of the crystal-melt interface kinetic coefficients measured from the non-equilibrium molecular-dynamics simulations.

Learning brain representation using recurrent Wasserstein generative adversarial net.

BACKGROUND AND OBJECTIVE: To understand brain cognition and disorders, modeling the mapping between mind and brain has been of great interest to the neuroscience community. The key is the brain representation, including functional brain networks (FBN) and their corresponding temporal features. Recently, it has been proven that deep learning models have superb representation power on functional magnetic resonance imaging (fMRI) over traditional machine learning methods. However, due to the lack of high-quality data and labels, deep learning models tend to suffer from overfitting in the training process. METHODS: In this work, we applied a recurrent Wasserstein generative adversarial net (RWGAN) to learn brain representation from volumetric fMRI data. Generative adversarial net (GAN) is widely used in natural image generation and is able to capture the distribution of the input data, which enables the extraction of generalized features from fMRI and thus relieves the overfitting issue. The recurrent layers in RWGAN are designed to better model the local temporal features of the fMRI time series. The discriminator of RWGAN works as a deep feature extractor. With LASSO regression, the RWGAN model can decompose the fMRI data into temporal features and spatial features (FBNs). Furthermore, the generator of RWGAN can generate high-quality new data for fMRI augmentation. RESULTS: The experimental results on seven tasks from the HCP dataset showed that the RWGAN can learn meaningful and interpretable temporal features and FBNs, compared to HCP task designs and general linear model (GLM) derived networks. Besides, the results on different training datasets showed that the RWGAN performed better on small datasets than other deep learning models. Moreover, we used the generator of RWGAN to yield fake subjects. The result showed that the fake data can also be used to learn meaningful representation compared to those learned from real data. CONCLUSIONS: To our best knowledge, this work is among the earliest attempts of applying generative deep learning for modeling fMRI data. The proposed RWGAN offers a novel methodology for learning brain representation from fMRI, and it can generate high-quality fake data for the potential use of fMRI data augmentation.

A New Nonlinear Photothermal Iterative Theory for Port-Wine Stain Detection.

The development of appropriate photothermal detection of skin diseases to meet complex clinical demands is an urgent challenge for the prevention and therapy of skin cancer. An extensive body of literature has ignored all high-order harmonics above the second order and their influences on low-order harmonics. In this paper, a new iterative numerical method is developed for solving the nonlinear thermal diffusion equation to improve nonlinear photothermal detection for the noninvasive assessment of the thickness of port-wine stain (PWS). First, based on the anatomical and structural properties of skin tissue of PWS, a nonlinear theoretical model for photothermal detection is established. Second, a corresponding nonlinear thermal diffusion equation is solved by using the new iterative numerical method and taking into account harmonics above the second-order and their effects on lower-order harmonics. Finally, the thickness and excitation light intensity of PWS samples are numerically simulated. The simulation results show that the numerical solution converges fasterand the physical meaning of the solution is clearerwith the new method than with the traditional perturbation method. The rate of change in each harmonic with the sample thickness for the new method is higher than that for the conventional perturbation method, suggesting that the proposed numerical method may provide greater detection sensitivity. The results of the study provide a theoretical basis for the clinical treatment of PWS.

Understanding the kinetic anisotropy of the soft-sphere bcc crystal-melt interfaces.

By employing the non-equilibrium molecular dynamics (MD) simulations and the time-dependent Ginzburg-Landau (TDGL) theory for the solidification kinetics, we predict the kinetic coefficients for the bcc(100), (110), and (111) CMIs of the soft-spheres, which are modeled with the inverse-power repulsive potential, and compare with the previous reported data of the bcc Fe system. We confirm a universal-like behavior of the spatial integrations of the (density wave amplitudes) Ginzburg-Landau order parameter square-gradient for the bcc CMI systems. The TDGL predictions of the kinetic anisotropies for bcc soft-sphere and bcc Fe CMI systems are identical; both agree well with the MD measurement for the soft-sphere system but differ strongly with the MD measurement for the Fe system. This finding implies that the current TDGL theory reflects a preference of presenting the generic anisotropy relationship due to the interfacial particle packings but lacks the contribution parameter which addresses the specificities in the kinetic anisotropies owing to the particle-particle interactions. A hypothesis that the density relaxation times for the interface melt phases to be anisotropic and material-dependent is then proposed.

Is the transperineal ultrasonography approach effective for the diagnosis of rectocele?

BACKGROUND: Transperineal ultrasonography has been used as a diagnostic imaging modality for rectocele for many years. However, the consistency of ultrasonography and defecography in evaluating the severity of rectocele was not satisfactory. This study aimed to evaluate the agreement in the measurement of rectocele parameters between the two methods in different positions and provide clinical implications for the diagnosis of rectocele. METHODS: In this pilot study, participants were recruited in an outpatient clinic of a tertiary hospital between December 2017 and December 2019. All participants separately underwent defecation proctography at sitting and squatting positions, and undertook transperineal ultrasonography at left lateral, sitting, and squatting positions. The consistency of ultrasonography and defecography was evaluated. RESULTS: Thirty female volunteers with rectocele were included in this study. The degree of anorectal angle was significantly larger at rest and during contraction, maximal Valsalva, and evacuation; the depth of the rectocele was significantly deeper during maximal Valsalva and evacuation; and the length of the perineum descending was significantly longer during contraction and maximal Valsalva in using squatting position compared to the sitting position when performing the defecation proctography. The degree of anorectal angle, the depth of rectocele, the area of levator hiatus, and the volume of the rectocele were significantly different in using squatting, sitting, and left lateral positions when performing the transperineal ultrasonography. Bland-Altman semi-quantitative plots showed good consistency in the measurement of the anorectal angle and the depth of the rectocele between proctography and ultrasonography in both sitting and squatting positions. CONCLUSIONS: The findings of our study may be considered as the preliminary evidence to support the use of transperineal ultrasonography with sitting and squatting positions as the imaging test of choice for evaluating patients with rectocele.

Coaxial Electrohydrodynamic Bioprinting of Pre-vascularized Cell-laden Constructs for Tissue Engineering.

Recapitulating the vascular networks that maintain the delivery of nutrition, oxygen, and byproducts for the living cells within the three-dimensional (3D) tissue constructs is a challenging issue in the tissue-engineering area. Here, a novel coaxial electrohydrodynamic (EHD) bioprinting strategy is presented to fabricate thick pre-vascularized cell-laden constructs. The alginate and collagen/calcium chloride solution were utilized as the outer-layer and inner-layer bioink, respectively, in the coaxial printing nozzle to produce the core-sheath hydrogel filaments. The effect of process parameters (the feeding rate of alginate and collagen and the moving speed of the printing stage) on the size of core and sheath lines within the printed filaments was investigated. The core-sheath filaments were printed in the predefined pattern to fabricate lattice hydrogel with perfusable lumen structures. Endothelialized lumen structures were fabricated by culturing the core-sheath filaments with endothelial cells laden in the core collagen hydrogel. Multilayer core-sheath filaments were successfully printed into 3D porous hydrogel constructs with a thickness of more than 3 mm. Finally, 3D pre-vascularized cardiac constructs were successfully generated, indicating the efficacy of our strategy to engineer living tissues with complex vascular structures.

Modeling and augmenting of fMRI data using deep recurrent variational auto-encoder.

Objective. Recently, deep learning models have been successfully applied in functional magnetic resonance imaging (fMRI) modeling and associated applications. However, there still exist at least two challenges. Firstly, due to the lack of sufficient data, deep learning models tend to suffer from overfitting in the training process. Secondly, it is still challenging to model the temporal dynamics from fMRI, due to that the brain state is continuously changing over scan time. In addition, existing methods rarely studied and applied fMRI data augmentation.Approach. In this work, we construct a deep recurrent variational auto-encoder (DRVAE) that combined variational auto-encoder and recurrent neural network, aiming to address all of the above mentioned challenges. The encoder of DRVAE can extract more generalized temporal features from assumed Gaussian distribution of input data, and the decoder of DRVAE can generate new data to increase training samples and thus partially relieve the overfitting issue. The recurrent layers in DRVAE are designed to effectively model the temporal dynamics of functional brain activities. LASSO (least absolute shrinkage and selection operator) regression is applied on the temporal features and input fMRI data to estimate the corresponding spatial networks.Main results. Extensive experimental results on seven tasks from HCP dataset showed that the DRVAE and LASSO framework can learn meaningful temporal patterns and spatial networks from both real data and generated data. The results on group-wise data and single subject suggest that the brain activities may follow certain distribution. Moreover, we applied DRVAE on four resting state fMRI datasets from ADHD-200 for data augmentation, and the results showed that the classification performances on augmented datasets have been considerably improved.Significance. The proposed method can not only derive meaningful temporal features and spatial networks from fMRI, but also generate high-quality new data for fMRI data augmentation and associated applications.