An Accelerated Approach on Adaptive Gradient Neural Network for Solving Time-Dependent Linear Equations: A State-Triggered Perspective.

To improve the acceleration performance, a hybrid state-triggered discretization (HSTD) is proposed for the adaptive gradient neural network (AGNN) for solving time-dependent linear equations (TDLEs). Unlike the existing approaches that use an activation function or a time-varying coefficient for acceleration, the proposed HSTD is uniquely designed from a control theory perspective. It comprises two essential components: adaptive sampling interval state-triggered discretization (ASISTD) and adaptive coefficient state-triggered discretization (ACSTD). The former addresses the gap in acceleration methods related to the variable sampling period, while the latter considers the underlying evolutionary dynamics of the Lyapunov function to determine coefficients greedily. Finally, compared with commonly used discretization methods, the acceleration performance and computational advantages of the proposed HSTD are substantiated by the numerical simulations and applications to robotics.

Enhanced axon outgrowth of spinal motor neurons in co-culturing with dorsal root ganglions antagonizes the growth inhibitory environment.

Introduction: Forming a bridge made of functional axons to span the lesion is essential to reconstruct the motor circuitry following spinal cord injury (SCI). Dorsal root ganglion (DRG) axons are robust in axon growth and have been proved to facilitate the growth of cortical neurons in a process of axon-facilitated axon regeneration. However, whether DRG transplantation affects the axon outgrowth of spinal motor neurons (SMNs) that play crucial roles in motor circuitry remains unclear. Methods: We investigated the axonal growth patterns of co-cultured DRGs and SMN aggregates (SMNAs) taking advantage of a well-designed 3D-printed in vitro system. Chondroitin sulphate proteoglycans (CSPG) induced inhibitory matrix was introduced to imitate the inhibitory environment following SCI. Axonal lengths of DRG, SMNA or DRG & SMNA cultured on the permissive or CSPG induced inhibitory matrix were measured and compared. Results: Our results indicated that under the guidance of full axonal connection generated from two opposing populations of DRGs, SMNA axons were growth-enhanced and elongated along the DRG axon bridge to distances that they could not otherwise reach. Quantitatively, the co-culture increased the SMNA axonal length by 32.1 %. Moreover, the CSPG matrix reduced the axonal length of DRGs and SMNAs by 46.2 % and 17.7 %, respectively. This inhibitory effect was antagonized by the co-culture of DRGs and SMNAs. Especially for SMNAs, they extended the axons across the CSPG-coating matrix, reached the lengths close to those of SMNAs cultured on the permissive matrix alone. Conclusions: This study deepens our understanding of axon-facilitated reconstruction of the motor circuitry. Moreover, the results support SCI treatment utilizing the enhanced outgrowth of axons to restore functional connectivity in SCI patients.

Mendelian randomization analysis suggests no associations of human herpes viruses with amyotrophic lateral sclerosis.

Background: The causal associations between infections with human herpes viruses (HHVs) and amyotrophic lateral sclerosis (ALS) has been disputed. This study investigated the causal associations between herpes simplex virus (HSV), varicella-zoster virus (VZV), Epstein-Barr virus (EBV), cytomegalovirus (CMV), HHV-6, and HHV-7 infections and ALS through a bidirectional Mendelian randomization (MR) method. Methods: The genome-wide association studies (GWAS) database were analyzed by inverse variance weighted (IVW), MR-Egger, weighted median, simple mode, and weighted mode methods. MR-Egger intercept test, MR-PRESSO test, Cochran's Q test, funnel plots, and leaveone-out analysis were used to verify the validity and robustness of the MR results. Results: In the forward MR analysis of the IVW, genetically predicted HSV infections [odds ratio (OR) = 0.9917; 95% confidence interval (CI): 0.9685-1.0154; p = 0.4886], HSV keratitis and keratoconjunctivitis (OR = 0.9897; 95% CI: 0.9739-1.0059; p = 0.2107), anogenital HSV infection (OR = 1.0062; 95% CI: 0.9826-1.0304; p = 0.6081), VZV IgG (OR = 1.0003; 95% CI: 0.9849-1.0160; p = 0.9659), EBV IgG (OR = 0.9509; 95% CI: 0.8879-1.0183; p = 0.1497), CMV (OR = 0.9481; 95% CI: 0.8680-1.0357; p = 0.2374), HHV-6 IgG (OR = 0.9884; 95% CI: 0.9486-1.0298; p = 0.5765) and HHV-7 IgG (OR = 0.9991; 95% CI: 0.9693-1.0299; p = 0.9557) were not causally associated with ALS. The reverse MR analysis of the IVW revealed comparable findings, indicating no link between HHVs infections and ALS. The reliability and validity of the findings were verified by the sensitivity analysis. Conclusion: According to the MR study, there is no evidence of causal associations between genetically predicted HHVs (HSV, VZV, EBV, CMV, HHV-6, and HHV-7) and ALS.

Comparison of HAV and HCV infections in vivo and in vitro reveals distinct patterns of innate immune evasion and activation.

BACKGROUND & AIMS: Hepatitis A virus (HAV) infections are considered not to trigger innate immunity in vivo, in contrast to hepatitis C virus (HCV). This lack of induction has been imputed to strong interference by HAV proteases 3CD and 3ABC. We aimed to elucidate the mechanisms of immune activation and counteraction by HAV and HCV in vivo and in vitro. METHODS: Albumin-urokinase-type plasminogen activator/severe combined immunodeficiency (Alb/uPA-SCID) mice with humanised livers were infected with HAV and HCV. Hepatic cell culture models were used to assess HAV and HCV sensing by Toll-like receptor 3 and retinoic acid-inducible gene I/melanoma differentiation-associated protein 5 (RIG-I/MDA5), respectively. Cleavage of the adaptor proteins TIR-domain-containing adapter-inducing interferon-ß (TRIF) and mitochondrial antiviral-signalling protein (MAVS) was analysed by transient and stable expression of HAV and HCV proteases and virus infection. RESULTS: We detected similar levels of interferon-stimulated gene induction in hepatocytes of HAV- and HCV-infected mice with humanised liver. In cell culture, HAV induced interferon-stimulated genes exclusively upon MDA5 sensing and depended on LGP2 (laboratory of genetics and physiology 2). TRIF and MAVS were only partially cleaved by HAV 3ABC and 3CD, not sufficiently to abrogate signalling. In contrast, HCV NS3-4A efficiently degraded MAVS, as previously reported, whereas TRIF cleavage was not detected. CONCLUSIONS: HAV induces an innate immune response in hepatocytes via MDA5/LGP2, with limited control of both pathways by proteolytic cleavage. HCV activates Toll-like receptor 3 and lacks TRIF cleavage, suggesting that this pathway mainly contributes to HCV-induced antiviral responses in hepatocytes. Our results shed new light on the induction of innate immunity and counteraction by HAV and HCV. IMPACT AND IMPLICATIONS: Understanding the mechanisms that determine the differential outcomes of HAV and HCV infections is crucial for the development of effective therapies. Our study provides insights into the interplay between these viruses and the host innate immune response in vitro and in vivo, shedding light on previously controversial or only partially investigated aspects. This knowledge could tailor the development of new strategies to combat HCV persistence, as well as improve our understanding of the factors underlying successful HAV clearance.

A zeroing neural dynamics based acceleration optimization approach for optimizers in deep neural networks.

The first-order optimizers in deep neural networks (DNN) are of pivotal essence for a concrete loss function to reach the local minimum or global one on the loss surface within convergence time. However, each optimizer possesses its own superiority and virtue when encountering a specific application scene and environment. In addition, the existing modified optimizers mostly emphasize a given optimizer without any transfer property. In this paper, a zeroing neural dynamics (ZND) based optimization approach for the first-order optimizers is proposed, which can assist ZND via the activation function to expedite the process of solving gradient information, with lower loss and higher accuracy. To the best of our knowledge, it is the first work to integrate the ZND in control domain with the first-order optimizers in DNN. This generic work is an optimization method for the most commonly-used first-order optimizers to handle different application scenes, rather than developing a brand-new algorithm besides the existing optimizers or their modifications. Furthermore, mathematic derivations concerning the gradient information transformation of the ZND are systematically provided. Finally, comparison experiments are implemented, which demonstrates the effectiveness of the proposed approach with different loss functions and network frameworks on the Reuters, CIFAR, and MNIST data sets.

Autonomous navigation of a magnetic colonoscope using force sensing and a heuristic search algorithm.

This paper presents an autonomous navigation system for cost-effective magnetic-assisted colonoscopy, employing force-based sensors, an actuator, a proportional-integrator controller and a real-time heuristic searching method. The force sensing system uses load cells installed between the robotic arm and external permanent magnets to derive attractive force data as the basis for real-time surgical safety monitoring and tracking information to navigate the disposable magnetic colonoscope. The average tracking accuracy on magnetic field navigator (MFN) platform in x-axis and y-axis are 1.14 ± 0.59 mm and 1.61 ± 0.45 mm, respectively, presented in mean error ± standard deviation. The average detectable radius of the tracking system is 15 cm. Three simulations of path planning algorithms are presented and the learning real-time A* (LRTA*) algorithm with our proposed directional heuristic evaluation design has the best performance. It takes 75 steps to complete the traveling in unknown synthetic colon map. By integrating the force-based sensing technology and LRTA* path planning algorithm, the average time required to complete autonomous navigation of a highly realistic colonoscopy training model on the MFN platform is 15 min 38 s and the intubation rate is 83.33%. All autonomous navigation experiments are completed without intervention by the operator.

Automatic lumen detection and magnetic alignment control for magnetic-assisted capsule colonoscope system optimization.

We developed a magnetic-assisted capsule colonoscope system with integration of computer vision-based object detection and an alignment control scheme. Two convolutional neural network models A and B for lumen identification were trained on an endoscopic dataset of 9080 images. In the lumen alignment experiment, models C and D used a simulated dataset of 8414 images. The models were evaluated using validation indexes for recall (R), precision (P), mean average precision (mAP), and F1 score. Predictive performance was evaluated with the area under the P-R curve. Adjustments of pitch and yaw angles and alignment control time were analyzed in the alignment experiment. Model D had the best predictive performance. Its R, P, mAP, and F1 score were 0.964, 0.961, 0.961, and 0.963, respectively, when the area of overlap/area of union was at 0.3. In the lumen alignment experiment, the mean degrees of adjustment for yaw and pitch in 160 trials were 21.70° and 13.78°, respectively. Mean alignment control time was 0.902 s. Finally, we compared the cecal intubation time between semi-automated and manual navigation in 20 trials. The average cecal intubation time of manual navigation and semi-automated navigation were 9 min 28.41 s and 7 min 23.61 s, respectively. The automatic lumen detection model, which was trained using a deep learning algorithm, demonstrated high performance in each validation index.

Aspirin Attenuates Hyperoxia-Induced Acute Respiratory Distress Syndrome (ARDS) by Suppressing Pulmonary Inflammation via the NF-κB Signaling Pathway.

Acute respiratory distress syndrome (ARDS) is a common destructive syndrome with high morbidity and mortality rates. Currently, few effective therapeutic interventions for ARDS are available. Clinical trials have shown that the effectiveness of aspirin is inconsistent. The contribution of platelets to the inflammatory response leading to the development of ARDS is increasingly recognized. The antiplatelet agent aspirin reportedly exerts a protective effect on acid- and hyperoxia-induced lung injury in murine models. Our previous study showed that pretreatment with aspirin exerts protective effects on hyperoxia-induced lung injury in mice. However, the mechanisms and therapeutic efficacy of aspirin in the posttreatment of hyperoxia-induced acute lung injury (ALI) remain unclear. In this study, we used a homozygous NF-κB-luciferase+/+ transgenic mouse model and treated mice with low-dose (25 µg/g) or high-dose (50 µg/g) aspirin at 0, 24, and 48 h after exposure to hyperoxia (inspired oxygen fraction (FiO2) > 95%). Hyperoxia-induced lung injury significantly increased the activation of NF-κB in the lung and increased the levels of macrophages infiltrating the lung and reactive oxygen species (ROS), increased the HO-1, NF-κB, TNF-α, IL-1ß, and IL-4 protein levels, and reduced the CC10, SPC, eNOS, Nrp-1, and IκBα protein levels in the lung tissue. Pulmonary edema and alveolar infiltration of neutrophils were also observed in the lung tissue of mice exposed to hyperoxia. However, in vivo imaging revealed that posttreatment with aspirin reduced luciferase expression, suggesting that aspirin might reduce NF-κB activation. Posttreatment with aspirin also reduced hyperoxia-induced increases in the numbers of lung macrophages, intracellular ROS levels, and the expression of TNF-α, IL-1ß, and IL-4; it also increased CC10, SPC and Nrp-1 levels compared with hyperoxia exposure alone. Lung histopathology also indicated that the aspirin posttreatment significantly reduced neutrophil infiltration and lung edema compared with hyperoxia exposure alone. Aspirin effectively induces an anti-inflammatory response in a model of hyperoxia-induced lung injury. Thus, aspirin may have potential as a novel treatment for hyperoxia-induced ALI.

Effects of dextrose prolotherapy on contusion-induced muscle injuries in mice.

Current treatment options for muscle injuries remain suboptimal and often result in delayed/incomplete recovery of damaged muscles. In this study, the effects of dextrose prolotherapy on inflammation and regeneration of skeletal muscles after a contusion injury were investigated. Mice were separated into five groups, including a normal control (NC), post-injury with no treatment (mass-drop injury, MDI), post-injury with 10% dextrose (MDI + 10% dextrose), post-injury with 20% dextrose (MDI + 20% dextrose), and post-injury with 30% dextrose (MDI + 30% dextrose). The gastrocnemius muscles of the mice were subjected to an MDI, and muscle samples were collected at 7 days post-injury. Results showed the serum creatine kinase (CK), blood urea nitrogen (BUN), creatinine (CREA), and low-density lipoprotein (LDH) of the MDI-alone group were significantly higher than those of the normal control group (p<0.05). However, levels of serum CK, BUN, CREA, and lactate dehydrogenase (LDH) significantly decreased with different concentrations of dextrose. In addition, dextrose suppressed the macrophage response (F4/80 protein decreased) and promoted muscle satellite cell regeneration (desmin protein increased). In conclusion, dextrose prolotherapy can effectively help repair muscles; therefore, it may be one of the methods for clinically treating muscle injuries.