Search

Remote Raman spectroscopy is a technique that can detect and identify different target molecules through Raman vibrational modes from a remote distance. However, the current remote Raman technique is restricted by poor detection sensitivity, and it is still extremely challenging for trace explosive detection. Here, in order to achieve trace explosive detection from a remote distance, we innovatively propose two enhanced Raman spectroscopy methods by using a plasmonic spray and a laser beam focusing/Raman signal collecting instrument. In brief, a facile convex lens can converge the laser beam and collect Raman scattering signals, and a plasmonic spray can be used for surface-enhanced Raman scattering. Under the combination of the above enhancement methods, we achieve remote Raman detection of a variety of trace explosives with a concentration of ∼1 μg/cm

Hongjun You,1 Qianqian Zhao,2 Mengya Dong1 1Department of Cardiovascular Medicine, Shaanxi Provincial People’s Hospital, Xi’an, 710068, Shaanxi, People’s Republic of China; 2Department of Clinical Immunology, The First Affiliated Hospital, Air Force Military Medical University, Xi’an, 710032, Shaanxi, People’s Republic of ChinaCorrespondence: Mengya Dong, Department of Cardiovascular Medicine, Shaanxi Provincial People’s Hospital, 256 West Youyi Road, Xi’an, Shaanxi, 710068, People’s Republic of China, Tel +86– 15802943974, Email 405895903@qq.comIntroduction: Accumulating evidences disclose that COVID-19, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has a marked effect on acute myocardial infarction (AMI). Nevertheless, the underlying pathophysiology correlation between the AMI and COVID-19 remains vague.Materials and Methods: Bioinformatics analyses of the altered transcriptional profiling of peripheral blood mononuclear cells (PBMCs) in patients with AMI and COVID-19 were implemented, including identification of differentially expressed genes and common genes between AMI and COVID-19, protein–protein interactions, Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway analyses, TF-genes and miRNA coregulatory networks, to explore their biological functions and potential roles in the pathogenesis of COVID-19-related AMI.Conclusion: Our bioinformatic analyses of gene expression profiling of PBMCs in patients with AMI and COVID-19 provide us with a unique view regarding underlying pathophysiology correlation between the two vital diseases.Keywords: AMI, SARS-CoV-2, differentially expressed genes, DEGs, bioinformatics, pathophysiology

Two-dimensional (2D) transition metal dichalcogenides (TMDCs) have been widely investigated for optoelectronic applications.

Currently, owing to the single-molecule-level sensitivity and highly informative spectroscopic characteristics, surface-enhanced Raman scattering (SERS) is regarded as the most direct and effective detection technique. However, SERS still faces several challenges in its practical applications, such as the complex matrix interferences, and low sensitivity to the molecules of intrinsic small cross-sections or weak affinity to the surface of metals. Here, we show an enrichment-typed sensing strategy with both excellent selectivity and ultrahigh detection sensitivity based on a powerful porous composite material, called mesoporous nanosponge. The nanosponge consists of porous β-cyclodextrin polymers immobilized with magnetic NPs, demonstrating remarkable capability of effective and fast removal of organic micropollutants, e.g., ~90% removal efficiency within ~1 min, and an enrichment factor up to ~103. By means of this current enrichment strategy, the limit of detection for typical organic pollutants can be significantly improved by 2~3 orders of magnitude. Consequently, the current enrichment strategy is proved to be applicable in a variety of fields for portable and fast detection, such as Raman and fluorescent sensing., The authors demonstrate a sensing approach for effectively monitoring organic micropollutants in water. They first use mesoporous nanosponge and magnetic nanoparticles for fast enrichment of micropollutants, and demonstrate SERS-based sensing with 2-3 orders of magnitude improvement in detection limits.

There is growing evidence of a strong relationship between COVID-19 and thrombosis. However, few bioinformatics-based analyses of critical genes and the mechanisms related to COVID-19 thrombosis existed. This study aimed to identify critical genes related to COVID-19 thrombosis by bioinformatic methods and explore the biological mechanisms and gene regulatory networks. The gene expression data were obtained from the Gene Expression Omnibus (GEO). Significant modular genes in GSE176480 were identified by weighted gene correlation network analysis and overlapped with differentially expressed genes by R package ‘DESeq2’ to obtain common genes. Functional enrichment analyses indicated that common genes were mainly enriched in biological processes such as platelet activation, signaling and aggregation, neutrophil degranulation and immune system and VEGFA-VEGFR2 signaling pathway et al. Finally, 16 genes (RPLP0, RPS4X, RPL13A, RPL13, RPL10, TPT1, PSMA7, ATP5D, AKT1, HIST1H2AC, HIST1H2BH, H3F3B, KDM6A, GATA3, ITGAM and RBMX) were identified as potential hub genes. Our study provides a new perspective to explore the pathogenesis and gene regulatory networks of thrombosis in COVID-19. It is worth highlighting that critical genes may be potential biomarkers and treatment targets of COVID-19 thrombosis for future study.

Accumulating evidence indicates the inflammatory state of rheumatoid arthritis (RA) predisposes to the acceleration of atherosclerosis (AS). Nevertheless, the potential mechanisms of accelerating AS in RA have not been fully elucidated. Our current study was to probe the problem via multi-microarray data analyses.The transcriptional profiling of synovial tissues from RA (GSE55235 and GSE55457) and that of atherosclerotic plaques from AS (GSE28829 and GSE41571) were downloaded from the Gene Expression Omnibus database. Bioinformatics analyses procedures included identifying common differentially expressed genes (DEGs), constructing protein-protein interaction network, key modules analysis and identifying hub genes, validating hub genes by using external datasets (GSE77298 and GSE163154), Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses, constructing transcription factor (TF)-miRNA coregulatory network and exploiting candidate drugs targeting hub genes.A total of 67 common DEGs were identified for downstream analyses. GO and KEGG analyses of these genes expounded a critical role of inflammatory mediators and reactions in the comorbidities. Sixteen hub genes were identified, and their functional analyses further highlighted a complicated inflammatory micro-environment and signaling pathways involving RA and AS. Six TFs and four miRNAs interacted with hub genes, and the candidate drugs targeting them were simvastatin, 5-azacytidine, bisindolylmaleimide, retinoic acid, and verteporfin, etc.Our comprehensive bioinformatics analyses provided a novel view regarding the potential pathogenesis of AS in RA. Furthermore, exploitation of candidate drugs might hold great promise in the future fight against the comorbidities.

In cardiac ischemia-reperfusion (I/R), autophagy of hyperactivated cardiomyocytes degrades normal proteins and organelles, destroys cells and causes irreversible cell death. The present study aimed to determine the molecular mechanism through which TNF-α-induced protein 8-like protein 2 (TIPE2) regulates cardiomyocyte apoptosis via autophagy in I/R. The results revealed that the number of apoptotic cells and the protein expression levels of TIPE2 in the heart tissue of I/R model mice were significantly increased.

Two-dimensional (2D) transition metal dichalcogenides (TMDCs) have been widely investigated for optoelectronic applications. Here, by employing the nanocasting method, molybdenum disulfide (MoS

To achieve ultrasensitive detection of trace targets through solution-based surface-enhanced Raman spectroscopy (SERS), direct adsorption of the target molecules on a SERS-active surface is vital. In this work, cetyltrimethylammonium bromide (CTAB)-capped gold nano-bipyramids (Au BPs) with different aspect ratios (ARs) are prepared and the surface is successfully modified by a simple ligand exchange method. Cysteamine-capped gold nano-bipyramids (cyst-Au BPs) are obtained by means of replacement of CTAB by cysteamine using Au-S covalent bonding and applied in the solution-based SERS detection of different pigment molecules, which always have weak affinity to the gold surface. The hydrogen bonding between the pigment molecule and cysteamine causes the aggregation of Au BPs to generate local electromagnetic field enhancement. The influence of the AR and concentration of Au BPs on SERS properties is investigated. The SERS detection of weak-affinity molecules to an extremely low limit shows that the cyst-Au BPs are highly sensitive compared to CTAB-capped Au BPs. The limit of detection (LOD) of allura red as low as 0.1 ppb and that of sunset yellow as low as 1 ppb show that the proposed strategy has many advantages due to its simplicity and fast and rapid detection for the sensitivity analysis of weak-affinity molecules.

PbO2–Co3O4–MnO2 electrodes, used in the electrowinning industry and in the degradation of organic pollutants, have demonstrated an elevated performance through macroscopic electrochemical measurements. However, few reports have investigated localized electrochemical performance, which plays an indispensable role in determining the essential reasons for the improvement of the modified material. In this study, the causes of the increase in electrochemical reactivity are unveiled from a micro perspective through scanning electrochemical microscopy (SECM), X-ray diffraction (XRD), Raman microscopy (Raman), and X-ray photoelectronic energy spectroscopy (XPS). The results show that the increase of electrochemical reactivity of the modified electrodes results from two factors: transformation of the microstructure and change in the intrinsic physicochemical properties. Constant-height scanning maps indicate that the electrochemical reactivity of the modified electrodes is higher than that of the PbO2 electrode on the whole and high-reactivity areas are orderly distributed, coinciding with the observations from SEM and XRD. Thus, one of the reasons for the improvement of the modified electrode performance is the refinement of the microscopic morphology. The other reason is the surge of the oxygen vacancy concentration on the surface of the coating, which is supported by XRD, Raman and XPS. This finding is detected by the probe approach curve (PAC), which can quantitatively characterize the electrochemical reactivity of a substrate. Heterogeneous charge transfer rate constants of the modified electrode are 4–5 times higher than that of the traditional PbO2 electrode. This research offers some insight into the electrochemical reactivity of modified PbO2 electrodes from a micro perspective.

We successfully prepared silver dendritic nanostructures with a well-defined and precisely tailored trunk and branches, as well as decorated nanoparticles, by a simple and controllable two-electrode electrochemical deposition system using different AgNO3 concentrations and deposition potentials. The prepared nanostructures are used as SERS substrates for the detection of crystal violet molecules with a minimum detection limit of 10−11 M. Finite-difference time-domain (FDTD) simulation is used to analyze the SERS activity of diverse silver dendritic nanostructures. The SERS and FDTD results revealed that the pine leaf-like structure displays excellent sensitivity and reproducibility compared to all other structures due to a large number of hotspot sites between branch–branch and branch–particle junctions. This article provides a good understanding of electromagnetic field enhancement and shows that various silver dendritic nanostructures could be potential candidates for SERS-based detection.

The stability, dispersity, and surface chemical properties of colloidal nanoparticles are crucial for the reliable and desired chemical sensing in various applications. Here, we report an effective strategy to engineer the surface properties of mesoporous Au nanoparticles (meso-Au NPs) via PVP ligand modification, template removal, and surface purification. Monodispersed 3D meso-Au NPs with well-defined sizes and shapes were obtained using a general soft-enveloping strategy. During surface modification, the addition of PVP ligands and the concentration of HF solutions play key roles in the stability, shape, and size distributions of ordered Au networks. In order to obtain an improved sensing performance, the morphologies of meso-Au NPs were optimized with smaller mesopore size, and NaBH4 solution was used to efficiently remove the adsorbed PVP ligands. Due to the characteristics of high-density porosities and large surface area, the purified meso-Au NPs could be a kind of promising plasmonic-enhanced nanomaterial and provide abundant "hot spots." Combined with the enrichment effect using a slippery liquid-infused porous surface, the lowest detection limits of crystal violet molecule could be down to 0.1 pM, demonstrating an excellent SERS sensitivity. Moreover, a realistic illegal drug containing aspirin could be sensitively detected with a limit of 2.8 × 10-6 M, showing great potential for practical molecular sensing and applications.

How to fabricate Au nanostructures conveniently on microstructured/nanostructured arrays surface with low cost has become a crucial and urgent challenge. In this study, we demonstrate hierarchical flowerlike Au nanostructures with rich nanothorns (HF-AuNTs) through one-step electrochemical deposition. The morphology of the HF-AuNTs is easily manipulated by controlling the applied potential or precursor solution concentration of electrodeposition. The as-prepared HF-AuNTs possessing unique local morphology of thin petals and dense thorns are further applied in the Si micropit arrays to acquire HF-AuNTs microarrays. As an initial detection, these HF-AuNTs microarrays exhibit a fascinating surface-enhanced Raman spectroscopy consistency (relative standard deviation is 7.17%) and sensitivity with the limitation of crystal violet reaching to 10-10 M, and Rhodamine 6G reaching to 10-11 M. The HF-AuNTs microarrays with well-defined shape and elaborate structure may be applicated in SERS substrates, superhydrophobic materials, and so on.

Detecting matter at a single-molecule level is the ultimate target in many branches of study. Nanosensors based on plasmonics have garnered significant interest owing to their ultrahigh sensitivity even at single-molecule level. However, currently, plasmonic-enhanced nanosensors have not achieved excellent performances in practical applications and their detection at femtomolar or attomolar concentrations remains highly challenging. Here we show a plasmonic sensing strategy, called buoyant plasmonic-particulate-based few-to-single particle-nanosensors. Large-sized floating particles combined with a slippery surface may prevent the coffee-ring effect and enhance the spatial enrichment capability of the analyte in plasmonic sensitive sites via the aggregation and lifting effect. Dimer and single particle-nanosensors demonstrate an enhanced surface-enhanced Raman spectroscopy (SERS) and a high fluorescence sensitivity with an enrichment factor up to an order of ∼104 and the limit of detection of CV molecules down to femto- or attomolar levels. The current buoyant particulate strategy can be exploited in a wide range of plasmonic enhanced sensing applications for a cost-effective, simple, fast, flexible, and portable detection., Plasmonic-enhanced nanosensors are limited in practical applications, as it remains challenging to detect molecules at low concentrations. Here, the authors introduce a buoyant particulate strategy in order to enrich analytes in the plasmonic hot spots.

Enriching and locating target analytes into specific "hot spots" are vital for ultrasensitive molecular identification and detection using plasmonic-based techniques. Inspired by mass transportation in lamp wicks, we develop an effective enrichment strategy for highly diluted analytes in which analytes and Au nanoparticles are transported via a solution microflow under the capillarity driving force of glass fiber papers to a heated region. After evaporation, a large volume of a solution containing analytes and Au nanoparticles is condensed into a very limited area, and thus, analyte molecules are effectively enriched and located into surface-enhanced Raman scattering (SERS) hot spots. Using this enrichment strategy, the sensitivity and detection limits of SERS are remarkably improved. Detection levels of crystal violet and anthracene are down to 10-16 and 10-10 M, respectively. This enrichment strategy is very robust and easy to implement, and it can potentially be exploited in various plasmonic-based molecular detection and identification techniques.

Layered two-dimensional (2D) materials and their heterostructures possess excellent optoelectronic properties due to their unique planar features. However, planar structures can only selectively support the fundamental optical modes, which is averse to fully exploit the potential of the 2D materials. Here, a novel type of tungsten disulfide (WS2) nanoparticle (NP) with a uniform size and morphology and highly ordered WS2 supercrystals (SCs) are synthesized by a nanocasting process using ordered mesoporous silica as a template. Due to the curved feature of individual nanostructures, their Raman signals show complex dependence behavior on the excitation wavelength, excitation power and temperature. Significantly, the silent phonon mode becomes Raman active due to the curvature of the interlaced WS2 layers. We believe that curved features will greatly enrich the optoelectronic applications of 2D materials.

Direct methanol fuel cells (DMFCs) have bright prospects because of their high energy density, simple structure, small fuel cartridge, instant recharging, and ease of storage and transport. Noble m...

The ability to engineer the surface ligands or adsorbed molecules on colloid nanocrystals (NCs) is important for various applications, as the physical and chemical properties are strongly affected by the surface chemistry. Here, we develop a facile and generalized ionic compound-mediated ligand-exchange strategy based on density functional theory calculations, in which the ionic compounds possess switchable bonding energy when they transfer between the ionized state and the non-ionized state, hence catalyzing the ligand-exchange process. By using an organic acid as the intermediate ligand, ligands such as oleylamine, butylamine, polyvinylpyrrolidone, and poly(vinyl alcohol) can be freely exchanged on the surface of Au NCs. Benefiting from this unique ligand-exchange strategy, the ligands with strong bonding energy can be replaced by weak ones, which is hard to realize in traditional ligand-exchange processes. The ionic compound-mediated ligand exchange is further utilized to improve the catalytic properties of Au NCs, facilitate the loading of nanoparticles on substrates, and tailor the growth of colloid NCs. These results indicate that the mechanism of switchable bonding energy can be significantly expanded to manipulate the surface property and functionalization of NCs that have applications in a wide range of chemical and biomedical fields.

The fabrication of flexible surface-enhanced Raman scattering (SERS) substrates with high sensitivity and signal consistency in a simple and low-cost way have been a crucial and urgent challenge. In this study, we demonstrated a flexible transparent three-dimensional (3D) ordered micro-hemisphere (MHS) array PDMS film with self-assembled Au nanoparticles (NPs) (an Au NP-MHS array film) as a surface-enhanced Raman scattering (SERS) platform for the in situ detection of pesticides in food. According to the finite difference time domain (FDTD) simulation results, the Au NP-MHS array film possesses high-density and ordered “hot spots” to enhance the electromagnetic field (EMF), so that the sensitivity and signal consistency are effectively improved. The optimized Au NP-MHS array film exhibits fascinating SERS sensitivity, with a limitation of crystal violet (CV) reaching 10−12 M as well as excellent consistency (RSD = 4.89% at 10−8 M). For in situ pesticide detection in fruits, this flexible film was pasted on the surfaces of tomatoes. The carbendazol and parathion were directly detected, and distinctly identified even at a low concentration (10−8 M). The flexible Au NP-MHS array film could be a potential candidate for the SERS substrate applied in the rapid and true inspection.

Uniform spherical Ag NPs with controllable size are synthesized in a simple one-step aqueous solution system. A unique digestive ripening process is observed using in situ UV-vis spectra, which performs key functions for the formation of ultra-uniform Ag NPs.

PbO

BACKGROUND The aim of the present study was to explore the predictive value of serum apurinic/apyrimidinic endonuclease 1/redox factor-1 (Ape1/Ref-1) combined with transforming growth factor β1 (TGF-β1) levels in the occurrence of radiation pneumonitis (RP) in patients with non-small cell lung cancer (NSCLS). METHODS Eighty-one patients with NSCLS who were admitted from August 2017 to July 2019 were enrolled in the present study. All patients were treated with concurrent radiotherapy and chemotherapy. Serum Ape1/Ref-1 and TGF-β1 levels were measured before treatment and 12 weeks after treatment. Patients with radiation-induced lung injury were assessed and divided into the RP group (lung injury ≥2) and non-RP (NRP) group (grade

Mitochondrial injury plays an essential role in the pathogenesis of diabetic cardiomyopathy (DCM). Previous studies demonstrated that rosmarinic acid (RA) treatment prevented high glucose-induced mitochondrial injury in vitro. However, whether RA can ameliorate cardiac function by preventing mitochondrial injury in DCM is unknown. The SIRT1/PGC-1α pathway has emerged as an important regulator of metabolic control and other mitochondrial functions. The present study was undertaken to determine the effects of RA on mitochondrial and cardiac function in DCM as well as the involvement of the SIRT1/PGC-1α pathway. Our results revealed that RA improved cardiac systolic and diastolic function and prevented mitochondrial injury in DCM, as shown by the reduced blood glucose and lipid levels, increased mitochondrial membrane potential levels, improved adenosine triphosphate synthesis, and inhibited apoptosis (P < 0.05). Moreover, RA upregulated the expression of SIRT1 and PGC-1α in DCM mice and high glucose-treated H9c2 cardiomyocytes (P < 0.05). Further mechanistic studies in H9c2 cardiomyocytes revealed that suppression of SIRT1 by Sh-SIRT1 counteracted the effects of RA on high glucose-induced abnormal metabolism of glucose and lipids, oxidative stress and apoptosis (P < 0.05). Taken together, these data indicate that RA prevented mitochondrial injury and cardiac dysfunction in DCM mice, and the SIRT1/PGC-1α pathway mediated the protective effects of RA.

Developing advanced sensing and detection technologies to effectively monitor organic micropollutants in water is under urgent demand in both scientific and industrial communities. Currently, owing to the ultrahigh sensitivity on the single-molecule level with highly informative spectra characteristics, SERS technique is regarded as the most direct and effective detection technique. However, some weakly adsorbed molecules, such as most of persistent organic pollutants, cannot exhibit strong SERS signals, which is a long-standing key challenge that has not been solved. Here, we show an enrichment-typed sensing strategy based on a powerful porous composite material, call mesoporous nanosponge. The nanosponge consists of magnetic nanoparticles immobilized porous β-cyclodextrin polymers, demonstrating remarkable capability of effective and fast removal of organic micropollutants, e.g. ~90% removal efficiency within ~1 min. With the anchoring of magnetic nanoparticles, the current new polymer adsorbent can be easily recycled from water and re-dispersed in ethanol so that the target molecules in the cavity of adsorbent is concentrated, with an enrichment factor up to ~103. By means of the current enrichment strategy, the limit of detection (LOD) of the typical organic pollutants can be significantly improved, i.e. increasing 2~3 orders of magnitude, compared with the detection without molecule enrichment protocol. Consequently, the current enrichment strategy is proved to be applicable in a variety of fields for portable and fast detection, such as Raman and fluorescent.

OBJECTIVE: To evaluate the effect of rosmarinic acid (RA) on mitophagy and hypertrophy of cardiomyocytes exposed to high glucose (HG). METHODS: Rat cardiomyocytes (H9c2) exposed to HG (25 mmol/L) were treated with 50 μmol/L RA or with both RA treatment and Parkin siRNA transfection, with the cells cultured in normal glucose (5.5 mmol/L) and HG as the controls. The expressions of PINK1, Parkin and LC3II/LC3I in the cells were detected by Western blotting. The formation of mitochondrial autophagosomes was observed by transmission electron microscope. Flow cytometry was employed to detect the level of reactive oxygen species (ROS) and apoptotic rate of the cells. The activities of respiratory chain complex enzymes were measured by spectrophotometry. Fluorescence enzyme labeling and (3)H-leucine labeling were used for determining the level of membrane potential and protein synthesis rate, respectively. The cell surface area was observed by light microscopy. RESULTS: RA treatment significantly increased the expression levels of PINK1, Parkin and LC3-II/I (P < 0.05), promoted the formation of mitochondrail autophagosome, inhibited the production of reactive oxygen species (P < 0.05), restored the activities of mitochondrial respiratory chain complex enzymes and mitochondrial membrane potential (P < 0.05), inhibited apoptosis (P < 0.05), and reduced the cell surface area and protein synthesis rate of H9c2 cells induced by HG exposure (P < 0.05). The protective effects of RA against HG-induced oxidative stress and cardiomyocyte hypertrophy was obviously blocked by inhibition of mitophagy mediated by transfection with Parkin siRNA (P < 0.05). CONCLUSION: RA can protect rat cardiomyocytes against oxidative stress injury and cardiomyocyte hypertrophy induced by HG by activating Parkin-mediated mitophagy.

High dependability computer system plays an important role in space application under space radiation environment. Registers in CPU (Central Processing Unit) and memory unit are the critical components of computer system security and the sensitive area of soft error. To assess the reliability of a safety-critical system, injecting random faults into these soft errors sensitive areas of the system to observe the behavior of the faulty system is a fast and effective method. In this paper, we present an efficient automated fault injection system FIG-QEMU (Fault Inject by GDB for QEMU) that simulates a variety of Single event effects, including Single event upset. The proposed system can assess not only the sensitivity of various space applications to soft errors, but also the robustness of the operating system. Our results show that 6.21% of the 5 million register faults and 3.53% of the 1.2 million memory faults will cause system crash respectively.

The absorption behavior between plasmonic nanostructures and a target molecule plays key roles in effective surface-enhanced Raman scattering (SERS) detection. However, for analytes with low surface affinity to the metallic surface, e.g., polycyclic aromatic hydrocarbons (PAHs), it remains challenging to observe the enhanced Raman signal. In this work, we reported a ratiometric SERS strategy for sensitive PAH detection through the surface functionalization of 3D ordered mesoporous Au nanoparticles (

Injecting faults to the system architecture layer and studying the upper neural network for fault tolerancehe is difficult and time-consuming. This paper proposes an automatic method covering time and space, which can inject faults into the processor on the Simics simulation platform, simulating soft errors, and then collect the time sequence data of the system architecture layer and the observed node data of visual convolutional neural networks program layer. At the same time, combined with the relevant standards, the GAN classifier is used to calibrate the different fault models after converting time sequence data into time sequence images. Finally, the Bayesian network is used to form the path of fault propagation from the architecture layer to the program layer and the result layer. After intensive fault injection into critical registers, the probability of neural network failure caused by soft errors is effectively stimulated.