Anomaly Detection Using Data Rate of Change on Medical Data

The identification and mitigation of anomaly data,characterized by deviations from normal patterns or singularities,stand as critical endeavors in modern technological landscapes,spanning domains such as Non-Fungible Tokens(NFTs),cyber-security,and the burgeoning metaverse.This paper presents a novel proposal aimed at refining anomaly detection methodologies,with a particular focus on continuous data streams.The essence of the proposed approach lies in analyzing the rate of change within such data streams,leveraging this dynamic aspect to discern anomalies with heightened precision and efficacy.Through empirical evaluation,our method demonstrates a marked improvement over existing techniques,showcasing more nuanced and sophisticated result values.Moreover,we envision a trajectory of continuous research and development,wherein iterative refinement and supplementation will tailor our approach to various anomaly detection scenarios,ensuring adaptability and robustness in real-world applications.

Video-Based Deception Detection with Non-Contact Heart Rate Monitoring and Multi-Modal Feature Selection

Deception detection plays a crucial role in criminal investigation.Videos contain a wealth of information regarding apparent and physiological changes in individuals,and thus can serve as an effective means of deception detection.In this paper,we investigate video-based deception detection considering both apparent visual features such as eye gaze,head pose and facial action unit(AU),and non-contact heart rate detected by remote photoplethysmography(rPPG)technique.Multiple wrapper-based feature selection methods combined with the K-nearest neighbor(KNN)and support vector machine(SVM)classifiers are employed to screen the most effective features for deception detection.We evaluate the performance of the proposed method on both a self-collected physiological-assisted visual deception detection(PV3D)dataset and a public bag-oflies(BOL)dataset.Experimental results demonstrate that the SVM classifier with symbiotic organisms search(SOS)feature selection yields the best overall performance,with an area under the curve(AUC)of 83.27%and accuracy(ACC)of 83.33%for PV3D,and an AUC of 71.18%and ACC of 70.33%for BOL.This demonstrates the stability and effectiveness of the proposed method in video-based deception detection tasks.

Trusted Encrypted Traffic Intrusion Detection Method Based on Federated Learning and Autoencoder

With the rapid development of the Internet,network security and data privacy are increasingly valued.Although classical Network Intrusion Detection System(NIDS)based on Deep Learning(DL)models can provide good detection accuracy,but collecting samples for centralized training brings the huge risk of data privacy leakage.Furthermore,the training of supervised deep learning models requires a large number of labeled samples,which is usually cumbersome.The“black-box”problem also makes the DL models of NIDS untrustworthy.In this paper,we propose a trusted Federated Learning(FL)Traffic IDS method called FL-TIDS to address the above-mentioned problems.In FL-TIDS,we design an unsupervised intrusion detection model based on autoencoders that alleviates the reliance on marked samples.At the same time,we use FL for model training to protect data privacy.In addition,we design an improved SHAP interpretable method based on chi-square test to perform interpretable analysis of the trained model.We conducted several experiments to evaluate the proposed FL-TIDS.We first determine experimentally the structure and the number of neurons of the unsupervised AE model.Secondly,we evaluated the proposed method using the UNSW-NB15 and CICIDS2017 datasets.The exper-imental results show that the unsupervised AE model has better performance than the other 7 intrusion detection models in terms of precision,recall and f1-score.Then,federated learning is used to train the intrusion detection model.The experimental results indicate that the model is more accurate than the local learning model.Finally,we use an improved SHAP explainability method based on Chi-square test to analyze the explainability.The analysis results show that the identification characteristics of the model are consistent with the attack characteristics,and the model is reliable.

Federated Network Intelligence Orchestration for Scalable and Automated FL-Based Anomaly Detection in B5G Networks

The management of network intelligence in Beyond 5G(B5G)networks encompasses the complex challenges of scalability,dynamicity,interoperability,privacy,and security.These are essential steps towards achieving the realization of truly ubiquitous Artificial Intelligence(AI)-based analytics,empowering seamless integration across the entire Continuum(Edge,Fog,Core,Cloud).This paper introduces a Federated Network Intelligence Orchestration approach aimed at scalable and automated Federated Learning(FL)-based anomaly detection in B5Gnetworks.By leveraging a horizontal Federated learning approach based on the FedAvg aggregation algorithm,which employs a deep autoencoder model trained on non-anomalous traffic samples to recognize normal behavior,the systemorchestrates network intelligence to detect and prevent cyber-attacks.Integrated into a B5G Zero-touch Service Management(ZSM)aligned Security Framework,the proposal utilizes multi-domain and multi-tenant orchestration to automate and scale the deployment of FL-agents and AI-based anomaly detectors,enhancing reaction capabilities against cyber-attacks.The proposed FL architecture can be dynamically deployed across the B5G Continuum,utilizing a hierarchy of Network Intelligence orchestrators for real-time anomaly and security threat handling.Implementation includes FL enforcement operations for interoperability and extensibility,enabling dynamic deployment,configuration,and reconfiguration on demand.Performance validation of the proposed solution was conducted through dynamic orchestration,FL,and real-time anomaly detection processes using a practical test environment.Analysis of key performance metrics,leveraging the 5G-NIDD dataset,demonstrates the system’s capability for automatic and near real-time handling of anomalies and attacks,including real-time network monitoring and countermeasure implementation for mitigation.

Improving Federated Learning through Abnormal Client Detection and Incentive

Data sharing and privacy protection are made possible by federated learning,which allows for continuous model parameter sharing between several clients and a central server.Multiple reliable and high-quality clients must participate in practical applications for the federated learning global model to be accurate,but because the clients are independent,the central server cannot fully control their behavior.The central server has no way of knowing the correctness of the model parameters provided by each client in this round,so clients may purposefully or unwittingly submit anomalous data,leading to abnormal behavior,such as becoming malicious attackers or defective clients.To reduce their negative consequences,it is crucial to quickly detect these abnormalities and incentivize them.In this paper,we propose a Federated Learning framework for Detecting and Incentivizing Abnormal Clients(FL-DIAC)to accomplish efficient and security federated learning.We build a detector that introduces an auto-encoder for anomaly detection and use it to perform anomaly identification and prevent the involvement of abnormal clients,in particular for the anomaly client detection problem.Among them,before the model parameters are input to the detector,we propose a Fourier transform-based anomaly data detectionmethod for dimensionality reduction in order to reduce the computational complexity.Additionally,we create a credit scorebased incentive structure to encourage clients to participate in training in order tomake clients actively participate.Three training models(CNN,MLP,and ResNet-18)and three datasets(MNIST,Fashion MNIST,and CIFAR-10)have been used in experiments.According to theoretical analysis and experimental findings,the FL-DIAC is superior to other federated learning schemes of the same type in terms of effectiveness.

Development of an Integrated Disposable Device for SARSCoV-2 Nucleic Acid Extraction and Detection

Objective To develop a highly sensitive and rapid nucleic acid detection method for the severe acute respiratory syndrome coronavirus 2(SARS-CoV-2).Methods We designed,developed,and manufactured an integrated disposable device for SARS-CoV-2 nucleic acid extraction and detection.The precision of the liquid transfer and temperature control was tested.A comparison between our device and a commercial kit for SARS-Cov-2 nucleic acid extraction was performed using real-time fluorescence reverse transcription polymerase chain reaction(RT-PCR).The entire process,from SARS-CoV-2 nucleic acid extraction to amplification,was evaluated.Results The precision of the syringe transfer volume was 19.2±1.9μL(set value was 20),32.2±1.6(set value was 30),and 57.2±3.5(set value was 60).Temperature control in the amplification tube was measured at 60.0±0.0℃(set value was 60)and 95.1±0.2℃(set value was 95)respectively.SARS-Cov-2 nucleic acid extraction yield through the device was 7.10×10^(6) copies/mL,while a commercial kit yielded 2.98×10^(6) copies/mL.The mean time to complete the entire assay,from SARS-CoV-2 nucleic acid extraction to amplification detection,was 36 min and 45 s.The detection limit for SARS-CoV-2 nucleic acid was 250 copies/mL.Conclusion The integrated disposable devices may be used for SARS-CoV-2 Point-of-Care test(POCT).

Development of an Integrated CMUTs-Based Resonant Biosensor for Label-Free Detection of DNA with Improved Selectivity by Ethylene-Glycol Alkanethiols

Gravimetric resonant-inspired biosensors have attracted increasing attention in industrial and point-ofcare applications,enabling label-free detection of biomarkers such as DNA and antibodies.Capacitive micromachined ultrasonic transducers(CMUTs)are promising tools for developing miniaturized highperformance biosensing complementary metal–oxide–silicon(CMOS)platforms.However,their operability is limited by inefficient functionalization,aggregation,crosstalk in the buffer,and the requirement for an external high-voltage(HV)power supply.In this study,we aimed to propose a CMUTs-based resonant biosensor integrated with a CMOS front–end interface coupled with ethylene–glycol alkanethiols to detect single-stranded DNA oligonucleotides with large specificity.The topography of the functionalized surface was characterized by energy-dispersive X-ray microanalysis.Improved selectivity for onchip hybridization was demonstrated by comparing complementary and non-complementary singlestranded DNA oligonucleotides using fluorescence imaging technology.The sensor array was further characterized using a five-element lumped equivalent model.The 4 mm^(2) application-specific integrated circuit chip was designed and developed through 0.18 lm HV bipolar-CMOS-double diffused metal–oxide–silicon(DMOS)technology(BCD)to generate on-chip 20 V HV boosting and to track feedback frequency under a standard 1.8 V supply,with a total power consumption of 3.8 mW in a continuous mode.The measured results indicated a detection sensitivity of 7.943×10^(-3) lmol·L^(-1)·Hz^(-1) over a concentration range of 1 to 100 lmol·L^(-1).In conclusion,the label-free biosensing of DNA under dry conditions was successfully demonstrated using a microfabricated CMUT array with a 2 MHz frequency on CMOS electronics with an internal HV supplier.Moreover,ethylene–glycol alkanethiols successfully deposited self-assembled monolayers on aluminum electrodes,which has never been attempted thus far on CMUTs,to enhance the selectivity of bio-functionalization.The findings of this study indicate the possibility of full-on-chip DNA biosensing with CMUTs.

Colonoscopy procedural volume increases adenoma and polyp detection rates in gastroenterology trainees

AIM To investigate changes in polyp detection throughout fellowship training, and estimate colonoscopy volume required to achieve the adenoma detection rate(ADRs) and polyp detection rate(PDRs) of attending gastroenterologists.METHODS We reviewed colonoscopies from July 1, 2009 to June 30, 2014. Fellows' procedural logs were used to retrieve colonoscopy procedural volumes, and these were treated as the time variable. Findings from screening colonoscopies were used to calculate colonoscopy outcomes for each fellow for the prior 50 colonoscopies at each time point. ADR and PDR were plotted against colonoscopy procedural volumes to produce individual longitudinal graphs. Repeated measures linear mixed effects models were used to study the change of ADR and PDR with increasing procedural volume.RESULTS During the study period, 12 fellows completed full three years of training and were included in the analysis. The average ADR and PDR were, respectively, 31.5% and 41.9% for all fellows, and 28.9% and 38.2% for attendings alone. There was a statistically significant increase in ADR with increasing procedural volume(1.8%/100 colonoscopies, P = 0.002). Similarly, PDR increased 2.8%/100 colonoscopies(P = 0.0001), while there was no significant change in advanced ADR(0.04%/100 colonoscopies, P = 0.92). The ADR increase was limited to the right side of the colon, while the PDR increased in both the right and left colon. The adenoma per colon and polyp per colon also increased throughout training. Fellows reached the attendings' ADR and PDR after 265 and 292 colonoscopies, respectively.CONCLUSION We found that the ADR and PDR increase with increasing colonoscopy volume throughout fellowship. Our findings support recent recommendations of ≥ 275 colonoscopies for colonoscopy credentialing.

A core-satellite self-assembled SERS aptasensor containing a“biological-silent region”Raman tag for the accurate and ultrasensitive detection of histamine

Herein,a novel interference-free surface-enhanced Raman spectroscopy(SERS)strategy based on magnetic nanoparticles(MNPs)and aptamer-driven assemblies was proposed for the ultrasensitive detection of histamine.A core-satellite SERS aptasensor was constructed by combining aptamer-decorated Fe_(3)O_(4)@Au MNPs(as the recognize probe for histamine)and complementary DNA-modified silver nanoparticles carrying 4-mercaptobenzonitrile(4-MBN)(Ag@4-MBN@Ag-c-DNA)as the SERS signal probe for the indirect detection of histamine.Under an applied magnetic field in the absence of histamine,the assembly gave an intense Raman signal at“Raman biological-silent”region due to 4-MBN.In the presence of histamine,the Ag@4-MBN@Ag-c-DNA SERS-tag was released from the Fe_(3)O_(4)@Au MNPs,thus decreasing the SERS signal.Under optimal conditions,an ultra-low limit of detection of 0.65×10^(-3)ng/mL and a linear range 10^(-2)-10^5 ng/mL on the SERS aptasensor were obtained.The histamine content in four food samples were analyzed using the SERS aptasensor,with the results consistent with those determined by high performance liquid chromatography.The present work highlights the merits of indirect strategies for the ultrasensitive and highly selective SERS detection of small biological molecules in complex matrices.

Customized Convolutional Neural Network for Accurate Detection of Deep Fake Images in Video Collections

The motivation for this study is that the quality of deep fakes is constantly improving,which leads to the need to develop new methods for their detection.The proposed Customized Convolutional Neural Network method involves extracting structured data from video frames using facial landmark detection,which is then used as input to the CNN.The customized Convolutional Neural Network method is the date augmented-based CNN model to generate‘fake data’or‘fake images’.This study was carried out using Python and its libraries.We used 242 films from the dataset gathered by the Deep Fake Detection Challenge,of which 199 were made up and the remaining 53 were real.Ten seconds were allotted for each video.There were 318 videos used in all,199 of which were fake and 119 of which were real.Our proposedmethod achieved a testing accuracy of 91.47%,loss of 0.342,and AUC score of 0.92,outperforming two alternative approaches,CNN and MLP-CNN.Furthermore,our method succeeded in greater accuracy than contemporary models such as XceptionNet,Meso-4,EfficientNet-BO,MesoInception-4,VGG-16,and DST-Net.The novelty of this investigation is the development of a new Convolutional Neural Network(CNN)learning model that can accurately detect deep fake face photos.

Association of trainee participation with adenoma and polyp detection rates

To investigate whether adenoma and polyp detection rates (ADR and PDR, respectively) in screening colonoscopies performed in the presence of fellows differ from those performed by attending physicians alone. METHODSWe performed a retrospective review of all patients who underwent a screening colonoscopy at Grady Memorial Hospital between July 1, 2009 and June 30, 2015. Patients with a history of colon polyps or cancer and those with poor colon preparation or failed cecal intubation were excluded from the analysis. Associations of fellowship training level with the ADR and PDR relative to attendings alone were assessed using unconditional multivariable logistic regression. Models were adjusted for sex, age, race, and colon preparation quality. RESULTSA total of 7503 colonoscopies met the inclusion criteria and were included in the analysis. The mean age of the study patients was 58.2 years; 63.1% were women and 88.2% were African American. The ADR was higher in the fellow participation group overall compared to that in the attending group: 34.5% vs 30.7% (P = 0.001), and for third year fellows it was 35.4% vs 30.7% (aOR = 1.23, 95%CI: 1.09-1.39). The higher ADR in the fellow participation group was evident for both the right and left side of the colon. For the PDR the corresponding figures were 44.5% vs 40.1% (P = 0.0003) and 45.7% vs 40.1% (aOR = 1.25, 95%CI: 1.12-1.41). The ADR and PDR increased with increasing fellow training level (P for trend < 0.05). CONCLUSIONThere is a stepwise increase in ADR and PDR across the years of gastroenterology training. Fellow participation is associated with higher adenoma and polyp detection.

Detection of Oxidants Such as Hydroxyl Radicals and Chlorine Electrogenerated on a BDD Electrode by Simple Methods

The aim of this work is to detect electrogenerated hydroxyl radicals and chlorine by simple and less expensive methods. Preparative electrolyses of perchloric acid (HClO4) and sodium chloride (NaCl) were performed on a boron-doped diamond (BDD) electrode. The hydroxyl radicals were quantified indirectly by assaying the samples from the HClO4 (0.1 M) electrolysis with a 10−4 M potassium permanganate solution. The investigations showed that the amount of hydroxyl radicals depends on the concentration of HClO4 and the current density. As for chlorine, a qualitative determination was carried out. A mixture of the electrolyte solution of HClO4 (0.1 M) + NaI (0.2 M) + 2 mL of hexane, taken in this order, leads to a purplish-pink coloration attesting to the presence of Cl2. The same test was carried out with NaBr and NaI giving pale and very pale pink colourations, respectively, showing that the intensity of the colouration depends on the strength of the oxidant present. In addition, oxidants were detected during the electrooxidation of metronidazole (MNZ). The results showed the participation of electrogenerated hydroxyl radicals. The generation of chlorine has also been proven. Furthermore, the degradation leads to a chemical oxygen demand (COD) removal rate of 83.48% and the process is diffusion-controlled.

Comparing Fine-Tuning, Zero and Few-Shot Strategies with Large Language Models in Hate Speech Detection in English

Large Language Models(LLMs)are increasingly demonstrating their ability to understand natural language and solve complex tasks,especially through text generation.One of the relevant capabilities is contextual learning,which involves the ability to receive instructions in natural language or task demonstrations to generate expected outputs for test instances without the need for additional training or gradient updates.In recent years,the popularity of social networking has provided a medium through which some users can engage in offensive and harmful online behavior.In this study,we investigate the ability of different LLMs,ranging from zero-shot and few-shot learning to fine-tuning.Our experiments show that LLMs can identify sexist and hateful online texts using zero-shot and few-shot approaches through information retrieval.Furthermore,it is found that the encoder-decoder model called Zephyr achieves the best results with the fine-tuning approach,scoring 86.811%on the Explainable Detection of Online Sexism(EDOS)test-set and 57.453%on the Multilingual Detection of Hate Speech Against Immigrants and Women in Twitter(HatEval)test-set.Finally,it is confirmed that the evaluated models perform well in hate text detection,as they beat the best result in the HatEval task leaderboard.The error analysis shows that contextual learning had difficulty distinguishing between types of hate speech and figurative language.However,the fine-tuned approach tends to produce many false positives.

Enhancing Log Anomaly Detection with Semantic Embedding and Integrated Neural Network Innovations

System logs,serving as a pivotal data source for performance monitoring and anomaly detection,play an indispensable role in assuring service stability and reliability.Despite this,the majority of existing log-based anomaly detection methodologies predominantly depend on the sequence or quantity attributes of logs,utilizing solely a single Recurrent Neural Network(RNN)and its variant sequence models for detection.These approaches have not thoroughly exploited the semantic information embedded in logs,exhibit limited adaptability to novel logs,and a single model struggles to fully unearth the potential features within the log sequence.Addressing these challenges,this article proposes a hybrid architecture based on amultiscale convolutional neural network,efficient channel attention and mogrifier gated recurrent unit networks(LogCEM),which amalgamates multiple neural network technologies.Capitalizing on the superior performance of robustly optimized BERT approach(RoBERTa)in the realm of natural language processing,we employ RoBERTa to extract the original word vectors from each word in the log template.In conjunction with the enhanced Smooth Inverse Frequency(SIF)algorithm,we generate more precise log sentence vectors,thereby achieving an in-depth representation of log semantics.Subsequently,these log vector sequences are fed into a hybrid neural network,which fuses 1D Multi-Scale Convolutional Neural Network(MSCNN),Efficient Channel Attention Mechanism(ECA),and Mogrifier Gated Recurrent Unit(GRU).This amalgamation enables themodel to concurrently capture the local and global dependencies of the log sequence and autonomously learn the significance of different log sequences,thereby markedly enhancing the efficacy of log anomaly detection.To validate the effectiveness of the LogCEM model,we conducted evaluations on two authoritative open-source datasets.The experimental results demonstrate that LogCEM not only exhibits excellent accuracy and robustness,but also outperforms the current mainstream log anomaly detection methods.

RRT Autonomous Detection Algorithm Based on Multiple Pilot Point Bias Strategy and Karto SLAM Algorithm

A Rapid-exploration Random Tree(RRT)autonomous detection algorithm based on the multi-guide-node deflection strategy and Karto Simultaneous Localization and Mapping(SLAM)algorithm was proposed to solve the problems of low efficiency of detecting frontier boundary points and drift distortion in the process of map building in the traditional RRT algorithm in the autonomous detection strategy of mobile robot.Firstly,an RRT global frontier boundary point detection algorithm based on the multi-guide-node deflection strategy was put forward,which introduces the reference value of guide nodes’deflection probability into the random sampling function so that the global search tree can detect frontier boundary points towards the guide nodes according to random probability.After that,a new autonomous detection algorithm for mobile robots was proposed by combining the graph optimization-based Karto SLAM algorithm with the previously improved RRT algorithm.The algorithm simulation platform based on the Gazebo platform was built.The simulation results show that compared with the traditional RRT algorithm,the proposed RRT autonomous detection algorithm can effectively reduce the time of autonomous detection,plan the length of detection trajectory under the condition of high average detection coverage,and complete the task of autonomous detection mapping more efficiently.Finally,with the help of the ROS-based mobile robot experimental platform,the performance of the proposed algorithm was verified in the real environment of different obstacles.The experimental results show that in the actual environment of simple and complex obstacles,the proposed RRT autonomous detection algorithm was superior to the traditional RRT autonomous detection algorithm in the time of detection,length of detection trajectory,and average coverage,thus improving the efficiency and accuracy of autonomous detection.

I-DCGAN and TOPSIS-IFP:A simulation generation model for radiographic flaw detection images in light alloy castings and an algorithm for quality evaluation of generated images

The intelligent detection technology driven by X-ray images and deep learning represents the forefront of advanced techniques and development trends in flaw detection and automated evaluation of light alloy castings.However,the efficacy of deep learning models hinges upon a substantial abundance of flaw samples.The existing research on X-ray image augmentation for flaw detection suffers from shortcomings such as poor diversity of flaw samples and low reliability of quality evaluation.To this end,a novel approach was put forward,which involves the creation of the Interpolation-Deep Convolutional Generative Adversarial Network(I-DCGAN)for flaw detection image generation and a comprehensive evaluation algorithm named TOPSIS-IFP.I-DCGAN enables the generation of high-resolution,diverse simulated images with multiple appearances,achieving an improvement in sample diversity and quality while maintaining a relatively lower computational complexity.TOPSIS-IFP facilitates multi-dimensional quality evaluation,including aspects such as diversity,authenticity,image distribution difference,and image distortion degree.The results indicate that the X-ray radiographic images of magnesium and aluminum alloy castings achieve optimal performance when trained up to the 800th and 600th epochs,respectively.The TOPSIS-IFP value reaches 78.7%and 73.8%similarity to the ideal solution,respectively.Compared to single index evaluation,the TOPSIS-IFP algorithm achieves higher-quality simulated images at the optimal training epoch.This approach successfully mitigates the issue of unreliable quality associated with single index evaluation.The image generation and comprehensive quality evaluation method developed in this paper provides a novel approach for image augmentation in flaw recognition,holding significant importance for enhancing the robustness of subsequent flaw recognition networks.

Model Agnostic Meta-Learning(MAML)-Based Ensemble Model for Accurate Detection of Wheat Diseases Using Vision Transformer and Graph Neural Networks

Wheat is a critical crop,extensively consumed worldwide,and its production enhancement is essential to meet escalating demand.The presence of diseases like stem rust,leaf rust,yellow rust,and tan spot significantly diminishes wheat yield,making the early and precise identification of these diseases vital for effective disease management.With advancements in deep learning algorithms,researchers have proposed many methods for the automated detection of disease pathogens;however,accurately detectingmultiple disease pathogens simultaneously remains a challenge.This challenge arises due to the scarcity of RGB images for multiple diseases,class imbalance in existing public datasets,and the difficulty in extracting features that discriminate between multiple classes of disease pathogens.In this research,a novel method is proposed based on Transfer Generative Adversarial Networks for augmenting existing data,thereby overcoming the problems of class imbalance and data scarcity.This study proposes a customized architecture of Vision Transformers(ViT),where the feature vector is obtained by concatenating features extracted from the custom ViT and Graph Neural Networks.This paper also proposes a Model AgnosticMeta Learning(MAML)based ensemble classifier for accurate classification.The proposedmodel,validated on public datasets for wheat disease pathogen classification,achieved a test accuracy of 99.20%and an F1-score of 97.95%.Compared with existing state-of-the-art methods,this proposed model outperforms in terms of accuracy,F1-score,and the number of disease pathogens detection.In future,more diseases can be included for detection along with some other modalities like pests and weed.

Defect Detection in Manufacturing: An Integrated Deep Learning Approach

This research investigates deep learning-based approach for defect detection in the steel production using Severstal steel dataset. The developed system integrates DenseNet121 for classification and DeepLabV3 for segmentation. DenseNet121 achieved high accuracy with defect classification as it achieved 92.34% accuracy during testing. This model significantly outperformed benchmark models like VGG16 and ResNet50, which achieved 72.59% and 92.01% accuracy, respectively. Similarly, for segmentation, DeepLabV3 showed high performance in localizing and categorizing defects, achieving a Dice coefficient of 84.21% during training and 69.77% during validation. The dataset includes steels which have four different types of defects and the DeepLab model was particularly effective with detection of Defect 4, with a Dice coefficient of 87.69% in testing. The model performs suboptimally in segmentation of Defect 1, achieving an accuracy of 64.81%. The overall system’s integration of classification and segmentation, alongside thresholding techniques, resulted in improved precision (92.31%) and reduced false positives. Overall, the proposed deep learning system achieved superior defect detection accuracy and reliability compared to existing models in the literature.

Endoscopic detection and diagnostic strategies for minute gastric cancer:A real-world observational study

BACKGROUND Minute gastric cancers(MGCs)have a favorable prognosis,but they are too small to be detected by endoscopy,with a maximum diameter≤5 mm.AIM To explore endoscopic detection and diagnostic strategies for MGCs.METHODS This was a real-world observational study.The endoscopic and clinicopathological parameters of 191 MGCs between January 2015 and December 2022 were retrospectively analyzed.Endoscopic discoverable opportunity and typical neoplastic features were emphatically reviewed.RESULTS All MGCs in our study were of a single pathological type,97.38%(186/191)of which were differentiated-type tumors.White light endoscopy(WLE)detected 84.29%(161/191)of MGCs,and the most common morphology of MGCs found by WLE was protruding.Narrow-band imaging(NBI)secondary observation detected 14.14%(27/191)of MGCs,and the most common morphology of MGCs found by NBI was flat.Another three MGCs were detected by indigo carmine third observation.If a well-demarcated border lesion exhibited a typical neoplastic color,such as yellowish-red or whitish under WLE and brownish under NBI,MGCs should be diagnosed.The proportion with high diagnostic confidence by magnifying endoscopy with NBI(ME-NBI)was significantly higher than the proportion with low diagnostic confidence and the only visible groups(94.19%>56.92%>32.50%,P<0.001).CONCLUSION WLE combined with NBI and indigo carmine are helpful for detection of MGCs.A clear demarcation line combined with a typical neoplastic color using nonmagnifying observation is sufficient for diagnosis of MGCs.MENBI improves the endoscopic diagnostic confidence of MGCs.

Detection of decline in estimated glomerular filtration rate in patients with type 2 diabetes by cystatin C-based equations

BACKGROUND Aging population is a significant issue in Viet Nam and across the globe.Elderly individuals are at higher risk of chronic kidney disease(CKD),especially those with diabetes.Several studies found that the estimated glomerular filtration rate(eGFR)determined using creatinine-based equations was not as accurate as that determined using cystatin C-based equations.Cystatin C-based equations may be beneficial in elderly patients with an age-associated decline in kidney function.Early determination of eGFR decline and associated factors would aid in appropriate interventions to improve kidney function in elderly patients with diabetes.AIM To determine the utility of cystatin C-based equations in early detection of eGFR decline and to explore factors associated with eGFR decline in elderly patients with diabetes.METHODS This cross-sectional study included 93 participants aged≥60 years evaluated in Can Tho University of Medicine and Pharmacy Hospital between October 2022 and July 2023,including 47 and 46 participants with and without diabetes respectively,according to the American Diabetes Association criteria for diabetes.The kappa coefficient,Student’s t,Mann-Whitney,χ2,Pearson’s correlation,multivariate logistic regression,and multiple linear regression analyses were employed.RESULTS The eGFRs were lower with the cystatin C-based equations than with the creatinine-based equations.Good agreement was found between the Modification of Diet in Renal Disease(MDRD)and CKD Epidemiology Collaboration(CKD-EPI)2021 creatinine-cystatin C equations(kappa=0.66).In the diabetes group,30%of the participants had low eGFR.Both plasma glucose and glycated hemoglobin were associated with an increased risk of eGFR decline(P<0.05)and negatively correlated with eGFR(P=0.001).By multivariate logistic regression,total cholesterol,and exercise were independently associated with low eGFR.By multiple linear regression,higher plasma glucose levels were correlated with lower eGFR(P=0.026,r=-0.366).CONCLUSION Cystatin C-based equations were superior in the early detection of a decline in eGFR,and the MDRD equation may be considered as an alternative to the CKD-EPI 2021 creatinine-cystatin C equation.Exercise,plasma glucose,and total cholesterol were independently associated with eGFR in patients with diabetes.