Three-Stage Transfer Learning with AlexNet50 for MRI Image Multi-Class Classification with Optimal Learning Rate

In radiology,magnetic resonance imaging(MRI)is an essential diagnostic tool that provides detailed images of a patient’s anatomical and physiological structures.MRI is particularly effective for detecting soft tissue anomalies.Traditionally,radiologists manually interpret these images,which can be labor-intensive and time-consuming due to the vast amount of data.To address this challenge,machine learning,and deep learning approaches can be utilized to improve the accuracy and efficiency of anomaly detection in MRI scans.This manuscript presents the use of the Deep AlexNet50 model for MRI classification with discriminative learning methods.There are three stages for learning;in the first stage,the whole dataset is used to learn the features.In the second stage,some layers of AlexNet50 are frozen with an augmented dataset,and in the third stage,AlexNet50 with an augmented dataset with the augmented dataset.This method used three publicly available MRI classification datasets:Harvard whole brain atlas(HWBA-dataset),the School of Biomedical Engineering of Southern Medical University(SMU-dataset),and The National Institute of Neuroscience and Hospitals brain MRI dataset(NINS-dataset)for analysis.Various hyperparameter optimizers like Adam,stochastic gradient descent(SGD),Root mean square propagation(RMS prop),Adamax,and AdamW have been used to compare the performance of the learning process.HWBA-dataset registers maximum classification performance.We evaluated the performance of the proposed classification model using several quantitative metrics,achieving an average accuracy of 98%.

Transfer RNA-derived small RNA serves as potential non-invasive diagnostic marker and a novel therapeutic target for acute pancreatitis

Transfer RNA(tRNA)-derived fragments,a new type of tRNA-derived small RNA(tsRNA),can be cleaved from tRNA by enzymes to regulate target gene expression at the transcriptional and translational levels.tsRNAs are not only degradation fragments but also have biological functions,including those in immune inflammation,metabolic disorders,and cell death.tsRNA dysregulation is closely associated with multiple diseases,including various cancers and acute pancreatitis(AP).AP is a common gastrointestinal disease,and its incidence increases annually.AP development is associated with tsRNAs,which regulate cell injury and induce inflammation,especially pyroptosis and ferroptosis.Notably,serum tRF36 has the potential to serve as a non-invasive diagnostic biomarker and leads to pancreatic acinar cell ferroptosis causing inflammation to promote AP.We show the characteristics of tsRNAs and their diagnostic value and function in AP,and discuss the potential opportunities and challenges of using tsRNAs in clinical applications and research.

Transfer learning for well logging formation evaluation using similarity weights

Machine learning has been widely applied in well logging formation evaluation studies.However,several challenges negatively impacted the generalization capabilities of machine learning models in practical imple-mentations,such as the mismatch of data domain between training and testing datasets,imbalances among sample categories,and inadequate representation of data model.These issues have led to substantial insufficient identification for reservoir and significant deviations in subsequent evaluations.To improve the transferability of machine learning models within limited sample sets,this study proposes a weight transfer learning framework based on the similarity of the labels.The similarity weighting method includes both hard weights and soft weights.By evaluating the similarity between test and training sets of logging data,the similarity results are used to estimate the weights of training samples,thereby optimizing the model learning process.We develop a double experts’network and a bidirectional gated neural network based on hierarchical attention and multi-head attention(BiGRU-MHSA)for well logs reconstruction and lithofacies classification tasks.Oil field data results for the shale strata in the Gulong area of the Songliao Basin of China indicate that the double experts’network model performs well in curve reconstruction tasks.However,it may not be effective in lithofacies classification tasks,while BiGRU-MHSA performs well in that area.In the study of constructing large-scale well logging processing and formation interpretation models,it is maybe more beneficial by employing different expert models for combined evaluations.In addition,although the improvement is limited,hard or soft weighting methods is better than unweighted(i.e.,average-weighted)in significantly different adjacent wells.The code and data are open and available for subsequent studies on other lithofacies layers.

Brachioradialis tendon transfer and palmaris longus tendon graft for thumb avulsion:A case report and review of literature

BACKGROUND Thumb replantation following complete traumatic avulsion requires complex techniques to restore function,especially in cases of avulsion at the level of the metacarpophalangeal joint(MCP I)and avulsion of the flexor pollicis longus(FPL)at the musculotendinous junction.Possible treatments include direct tendon suture or tendon transfer,most commonly from the ring finger.To optimize function and avoid donor finger complications,we performed thumb replantation with flexion restoration using brachioradialis(BR)tendon transfer with palmaris longus(PL)tendon graft.CASE SUMMARY A 20-year-old left-handed male was admitted for a complete traumatic left thumb amputation following an accident while sliding from the top of a handrail.The patient presented with skin and bone avulsion at the MCP I,avulsion of the FPL tendon at the musculotendinous junction(zone 5),avulsion of the extensor pollicis longus tendon(zone T3),and avulsion of the thumb’s collateral arteries and nerves.The patient was treated with two stage thumb repair.The first intervention consisted of thumb replantation with MCP I arthrodesis,resection of avulsed FPL tendon and implantation of a silicone tendon prosthesis.The second intervention consisted of PL tendon graft and BR tendon transfer.Follow-up at 10 months showed good outcomes with active interphalangeal flexion of 70°,grip strength of 45 kg,key pinch strength of 15 kg and two-point discrimination threshold of 4 mm.CONCLUSION Flexion restoration after complete thumb amputation with FPL avulsion at the musculotendinous junction can be achieved using BR tendon transfer with PL tendon graft.

Analytical method of load-transfer of single pile under expansive soil swelling

The elastic differential equations of load-transfer of single pile either with applied loads on pile-top or only under the soil swelling were established,respectively,based on the theory of pile-soil interaction and the shear-deformation method.The derivation of analytic solution to load-transfer for single pile in expansive soil could hereby be obtained by means of superposition principle under expansive soils swelling.The comparison of two engineering examples was made to prove the credibility of the suggested method.The analyzed results show that this analytic solution can achieve high precision with few parameters required,indicating its' simplicity and practicability in engineering application.The employed method can contribute to determining the greatest tension along pile shaft resulting from expansive soils swelling and provide reliable bases for engineering design.The method can be employed to obtain various distributive curves of axial force,settlements and skin friction along the pile shaft with the changes of active depth,vertical movements of the surface and loads of pile-top.

Modeling of Mass Transfer in Nonideal Multicomponent Mixture with Maxwell-Stefan Approach

The Intalox metal tower packing was used to simulate an industrial relevant extractive distillation column for purifying azeotropic multicomponent mixture.In order to explain the inconsistencies in the modeling of transfer process in nonideal multicomponent distillation column,a method was developed with equilibrium stage models(EQ)and non-equilibrium model(NEQ)incorporated with Maxwell-Stefan diffusion equations in the framework of AspenONE simulator.Dortmund Modified UNIFAC(UNIFAC-DMD)thermodynamic model was employed to estimate activity coefficients.In addition,to understand the reason for the diffusion against driving force and the different results by EQ and NEQ models,explicit investigations were made on diffusion coefficients, component Murphree efficiency and mass transfer coefficients.The results provide valuable information for basic design and applications associated with extractive distillation.

A Two-Layer Model for Superposed Electrified Maxwell Fluids in Presence of Heat Transfer

Based on a modified-Darcy-Maxwell model, two-dimensional, incompressible and heat transfer flow of two bounded layers, through electrified Maxwell fluids in porous media is performed. The driving force for the instability under an electric field, is an electrostatic force exerted on the free charges accumulated at the dividing interface. Normal mode analysis is considered to study the linear stability of the disturbances layers. The solutions of the linearized equations of motion with the boundary conditions lead to an implicit dispersion relation between the growth rate and wave number. These equations are parameterized by Weber number, Reynolds number, Marangoni number, dimensionless conductivities, and dimensionless electric potentials. The case of long waves interfaciaJ stability has been studied. The stability criteria are performed theoreticaily in which stability diagrams are obtained. In the limiting cases, some previously published results can be considered as particular cases of our results. It is found that the Reynolds number plays a destabilizing role in the stability criteria, while the damping influence is observed for the increasing of Marangoni number and Maxwell relaxation time.

Heat and Mass Transfer of Upper Convected Maxwell Fluid Flow with Variable Thermo-Physical Properties over a Horizontal Melting Surface

The objective of this article is to present the dynamics of an Upper Convected Maxwell (UCM) fluid flow with heat and mass transfer over a melting surface. The influence of melting heat transfer, thermal and solutal stratification are properly accounted for by modifying the classical boundary conditions of temperature and concentration respectively. It is assumed that the ratio of inertia forces to viscous forces is high enough for boundary layer approximation to be valid. The corresponding influence of exponential space dependent internal heat source on viscosity and thermal conductivity of UCM is properly considered. The dynamic viscosity and thermal conductivity of UCM are temperature dependent. Classical temperature dependent viscosity and thermal conductivity models were modified to suit the case of both melting heat transfer and thermal stratification. The governing non-linear partial differential equations describing the problem are reduced to a system of nonlinear ordinary differential equations using similarity transformations and completed the solution numerically using the Runge-Kutta method along with shooting technique. For accurate and correct analysis of the effect of variable viscosity on fluid flow in which (Tw or Tm) T∞ , the mathematical models of variable viscosity and thermal conductivity must be modified.

MHD flow and mass transfer of chemically reactive upper convected Maxwell fluid past porous surface

The magnetohydrodynamic （MHD） flow and mass transfer of an electrically conducting upper convected Maxwell （UCM） fluid at a porous surface are studied in the presence of a chemically reactive species. The governing nonlinear partial differential equations along with the appropriate boundary conditions are transformed into nonlinear ordinary differential equations and numerically solved by the Keller-box method. The effects of various physical parameters on the flow and mass transfer characteristics are graphically presented and discussed. It is observed that the order of the chemical reaction is to increase the thickness of the diffusion boundary layer. Also, the mass transfer rate strongly depends on the Schmidt number and the reaction rate parameter. Furthermore, available results in the literature are obtained as a special case.

IDS-INT:Intrusion detection system using transformer-based transfer learning for imbalanced network traffic

A network intrusion detection system is critical for cyber security against llegitimate attacks.In terms of feature perspectives,network traffic may include a variety of elements such as attack reference,attack type,a subcategory of attack,host information,malicious scripts,etc.In terms of network perspectives,network traffic may contain an imbalanced number of harmful attacks when compared to normal traffic.It is challenging to identify a specific attack due to complex features and data imbalance issues.To address these issues,this paper proposes an Intrusion Detection System using transformer-based transfer learning for Imbalanced Network Traffic(IDS-INT).IDS-INT uses transformer-based transfer learning to learn feature interactions in both network feature representation and imbalanced data.First,detailed information about each type of attack is gathered from network interaction descriptions,which include network nodes,attack type,reference,host information,etc.Second,the transformer-based transfer learning approach is developed to learn detailed feature representation using their semantic anchors.Third,the Synthetic Minority Oversampling Technique(SMOTE)is implemented to balance abnormal traffic and detect minority attacks.Fourth,the Convolution Neural Network(CNN)model is designed to extract deep features from the balanced network traffic.Finally,the hybrid approach of the CNN-Long Short-Term Memory(CNN-LSTM)model is developed to detect different types of attacks from the deep features.Detailed experiments are conducted to test the proposed approach using three standard datasets,i.e.,UNsWNB15,CIC-IDS2017,and NSL-KDD.An explainable AI approach is implemented to interpret the proposed method and develop a trustable model.

Causal Analysis Between Rice Growth and Cadmium Accumulation and Transfer under Arbuscular Mycorrhizal Inoculation

Cadmium(Cd)contamination in rice has been a serious threat to human health.To investigate the effects of arbuscular mycorrhizal fungi(AMF)on the Cd translocation in rice,a controlled pot experiment was conducted.The results indicated that AMF significantly increased rice biomass,with an increase of up to 40.0%,particularly in root biomass by up to 68.4%.Notably,the number of prominent rice individuals also increased,and their plasticity was enhanced following AMF inoculation.AMF led to an increase in the net photosynthetic rate and antioxidant enzyme activity of rice.In the AMF treatment group,the Cd concentration in the rice roots was significantly higher(19.1%‒68.0%)compared with that in the control group.Conversely,the Cd concentration in the rice seeds was lower in the AMF treatment group,indicating that AMF facilitated the sequestration of Cd in rice roots and reduced Cd accumulation in the seeds.Path coefficients varied across different treatments,suggesting that AMF inoculation reduced the direct impact of soil Cd concentration on the total Cd accumulation in seeds.The translocation of Cd was consistently associated with simultaneous growth dilution and compensatory accumulation as a result of mycorrhizal effects.Our study quantitatively analyzed this process through path analysis and clarified the causal relationship between rice growth and Cd transfer under the influence of AMF.

Effects of projectile parameters on the momentum transfer and projectile melting during hypervelocity impact

The effects of projectile/target impedance matching and projectile shape on energy,momentum transfer and projectile melting during collisions are investigated by numerical simulation.By comparing the computation results with the experimental results,the correctness of the calculation and the statistical method of momentum transfer coefficient is verified.Different shapes of aluminum,copper and heavy tungsten alloy projectiles striking aluminum,basalt,and pumice target for impacts up to 10 km/s are simulated.The influence mechanism of the shape of the projectile and projectile/target density on the momentum transfer was obtained.With an increase in projectile density and length-diameter ratio,the energy transfer time between the projectile and targets is prolonged.The projectile decelerates slowly,resulting in a larger cratering depth.The energy consumed by the projectile in the excavation stage increased,resulting in lower mass-velocity of ejecta and momentum transfer coefficient.The numerical simulation results demonstrated that for different projectile/target combinations,the higher the wave impedance of the projectile,the higher the initial phase transition velocity and the smaller the mass of phase transition.The results can provide theoretical guidance for kinetic impactor design and material selection.

Coherent optical frequency transfer via 972-km fiber link

We demonstrate coherent optical frequency dissemination over a distance of 972 km by cascading two spans where the phase noise is passively compensated for.Instead of employing a phase discriminator and a phase locking loop in the conventional active phase control scheme,the passive phase noise cancellation is realized by feeding double-trip beat-note frequency to the driver of the acoustic optical modulator at the local site.This passive scheme exhibits fine robustness and reliability,making it suitable for long-distance and noisy fiber links.An optical regeneration station is used in the link for signal amplification and cascaded transmission.The phase noise cancellation and transfer instability of the 972-km link is investigated,and transfer instability of 1.1×10^(-19)at 10^(4)s is achieved.This work provides a promising method for realizing optical frequency distribution over thousands of kilometers by using fiber links.

Generation of mitochondrial replacement monkeys by female pronucleus transfer

Mutations in mitochondrial DNA(mtDNA)are maternally inherited and have the potential to cause severe disorders.Mitochondrial replacement therapies,including spindle,polar body,and pronuclear transfers,are promising strategies for preventing the hereditary transmission of mtDNA diseases.While pronuclear transfer has been used to generate mitochondrial replacement mouse models and human embryos,its application in non-human primates has not been previously reported.In this study,we successfully generated four healthy cynomolgus monkeys(Macaca fascicularis)via female pronuclear transfer.These individuals all survived for more than two years and exhibited minimal mtDNA carryover(3.8%–6.7%),as well as relatively stable mtDNA heteroplasmy dynamics during development.The successful establishment of this nonhuman primate model highlights the considerable potential of pronuclear transfer in reducing the risk of inherited mtDNA diseases and provides a valuable preclinical research model for advancing mitochondrial replacement therapies in humans.

Label Recovery and Trajectory Designable Network for Transfer Fault Diagnosis of Machines With Incorrect Annotation

The success of deep transfer learning in fault diagnosis is attributed to the collection of high-quality labeled data from the source domain.However,in engineering scenarios,achieving such high-quality label annotation is difficult and expensive.The incorrect label annotation produces two negative effects:1)the complex decision boundary of diagnosis models lowers the generalization performance on the target domain,and2)the distribution of target domain samples becomes misaligned with the false-labeled samples.To overcome these negative effects,this article proposes a solution called the label recovery and trajectory designable network(LRTDN).LRTDN consists of three parts.First,a residual network with dual classifiers is to learn features from cross-domain samples.Second,an annotation check module is constructed to generate a label anomaly indicator that could modify the abnormal labels of false-labeled samples in the source domain.With the training of relabeled samples,the complexity of diagnosis model is reduced via semi-supervised learning.Third,the adaptation trajectories are designed for sample distributions across domains.This ensures that the target domain samples are only adapted with the pure-labeled samples.The LRTDN is verified by two case studies,in which the diagnosis knowledge of bearings is transferred across different working conditions as well as different yet related machines.The results show that LRTDN offers a high diagnosis accuracy even in the presence of incorrect annotation.

Robust Information Hiding Based on Neural Style Transfer with Artificial Intelligence

This paper proposes an artificial intelligence-based robust information hiding algorithm to address the issue of confidential information being susceptible to noise attacks during transmission.The algorithm we designed aims to mitigate the impact of various noise attacks on the integrity of secret information during transmission.The method we propose involves encoding secret images into stylized encrypted images and applies adversarial transfer to both the style and content features of the original and embedded data.This process effectively enhances the concealment and imperceptibility of confidential information,thereby improving the security of such information during transmission and reducing security risks.Furthermore,we have designed a specialized attack layer to simulate real-world attacks and common noise scenarios encountered in practical environments.Through adversarial training,the algorithm is strengthened to enhance its resilience against attacks and overall robustness,ensuring better protection against potential threats.Experimental results demonstrate that our proposed algorithm successfully enhances the concealment and unknowability of secret information while maintaining embedding capacity.Additionally,it ensures the quality and fidelity of the stego image.The method we propose not only improves the security and robustness of information hiding technology but also holds practical application value in protecting sensitive data and ensuring the invisibility of confidential information.

The regulation of ferrocene-based catalysts on heat transfer in highpressure combustion of ammonium perchlorate/hydroxyl-terminated polybutadiene/aluminum composite propellants

The regulation of the burning rate pressure exponent for the ammonium perchlorate/hydroxylterminated polybutadiene/aluminum(AP/HTPB/Al)composite propellants under high pressures is a crucial step for its application in high-pressure solid rocket motors.In this work,the combustion characteristics of AP/HTPB/Al composite propellants containing ferrocene-based catalysts were investigated,including the burning rate,thermal behavior,the local heat transfer,and temperature profile in the range of 7-28 MPa.The results showed that the exponent breaks were still observed in the propellants after the addition of positive catalysts(Ce-Fc-MOF),the burning rate inhibitor((Ferrocenylmethyl)trimethylammonium bromide,Fc Br)and the mixture of Fc Br/catocene(GFP).However,the characteristic pressure has increased,and the exponent decreased from 1.14 to 0.66,0.55,and 0.48 when the addition of Ce-FcMOF,Fc Br and Fc Br/GFP in the propellants.In addition,the temperature in the first decomposition stage was increased by 7.50℃ and 11.40℃ for the AP/Fc Br mixture and the AP/Fc Br/GFP mixture,respectively,compared to the pure AP.On the other hand,the temperature in the second decomposition stage decreased by 48.30℃ and 81.70℃ for AP/Fc Br and AP/Fc Br/GFP mixtures,respectively.It was also found that Fc Br might generate ammonia to cover the AP surface.In this case,a reaction between the methyl in Fc Br and perchloric acid caused more ammonia to appear at the AP surface,resulting in the suppression of ammonia desorption.In addition,the coarse AP particles on the quenched surface were of a concave shape relative to the binder matrix under low and high pressures when the catalysts were added.In the process,the decline at the AP/HTPB interface was only exhibited in the propellant with the addition of Ce-Fc-MOF.The ratio of the gas-phase temperature gradient of the propellants containing catalysts was reduced significantly below and above the characteristic pressure,rather than 3.6 times of the difference in the blank propellant.Overall,the obtained results demonstrated that the pressure exponent could be effectively regulated and controlled by adjusting the propellant local heat and mass transfer under high and low pressures.

Reliability analysis of slope stability by neural network,principal component analysis,and transfer learning techniques

The prediction of slope stability is considered as one of the critical concerns in geotechnical engineering.Conventional stochastic analysis with spatially variable slopes is time-consuming and highly computation-demanding.To assess the slope stability problems with a more desirable computational effort,many machine learning(ML)algorithms have been proposed.However,most ML-based techniques require that the training data must be in the same feature space and have the same distribution,and the model may need to be rebuilt when the spatial distribution changes.This paper presents a new ML-based algorithm,which combines the principal component analysis(PCA)-based neural network(NN)and transfer learning(TL)techniques(i.e.PCAeNNeTL)to conduct the stability analysis of slopes with different spatial distributions.The Monte Carlo coupled with finite element simulation is first conducted for data acquisition considering the spatial variability of cohesive strength or friction angle of soils from eight slopes with the same geometry.The PCA method is incorporated into the neural network algorithm(i.e.PCA-NN)to increase the computational efficiency by reducing the input variables.It is found that the PCA-NN algorithm performs well in improving the prediction of slope stability for a given slope in terms of the computational accuracy and computational effort when compared with the other two algorithms(i.e.NN and decision trees,DT).Furthermore,the PCAeNNeTL algorithm shows great potential in assessing the stability of slope even with fewer training data.

Impact of Magnetic Field on a Peristaltic Flow with Heat Transfer of a Fractional Maxwell Fluid in a Tube

Magnetic field and the fractional Maxwell fluids’impacts on peristaltic flows within a circular cylinder tube with heat transfer was evaluated while assuming that they are preset with a low-Reynolds number and a long wavelength.Utilizing,the fractional calculus method,the problem was solved analytically.It was deduced for temperature,axial velocity,tangential stress,and heat transfer coefficient.Many emerging parameters and their effects on the aspects of the flow were illustrated,and the outcomes were expressed via graphs.A special focus was dedicated to some criteria,such as the wave amplitude’s effect,Hartman and Grashof numbers,radius and relaxation–retardation ratios,and heat source,which were under discussions on the axial velocity,tangential stress,heat transfer,and temperature coefficients across one wavelength.Multiple graphs of physical interest were provided.The outcomes state that the effect of the criteria mentioned beforehand(the Hartman and Grashof numbers,wave amplitude,radius ratio,heat source,and relaxation–retardation ratio)were quite evident.

An Intelligent Diagnosis Method of the Working Conditions in Sucker-Rod Pump Wells Based on Convolutional Neural Networks and Transfer Learning

In recent years,deep learning models represented by convolutional neural networks have shown incomparable advantages in image recognition and have been widely used in various fields.In the diagnosis of sucker-rod pump working conditions,due to the lack of a large-scale dynamometer card data set,the advantages of a deep convolutional neural network are not well reflected,and its application is limited.Therefore,this paper proposes an intelligent diagnosis method of the working conditions in sucker-rod pump wells based on transfer learning,which is used to solve the problem of too few samples in a dynamometer card data set.Based on the dynamometer cards measured in oilfields,image classification and preprocessing are conducted,and a dynamometer card data set including 10 typical working conditions is created.On this basis,using a trained deep convolutional neural network learning model,model training and parameter optimization are conducted,and the learned deep dynamometer card features are transferred and applied so as to realize the intelligent diagnosis of dynamometer cards.The experimental results show that transfer learning is feasible,and the performance of the deep convolutional neural network is better than that of the shallow convolutional neural network and general fully connected neural network.The deep convolutional neural network can effectively and accurately diagnose the working conditions of sucker-rod pump wells and provide an effective method to solve the problem of few samples in dynamometer card data sets.