False Positive Research Articles

A dual-mode strategy for early detection of sugarcane pokkah boeng disease pathogen: A portable sensing device based on Cross-N DNA framework and MoS2@GDY

Sucrose, a common sugar primarily derived from sugarcane, is a crucial national strategic resource. However, its yield is significantly affected by various serious diseases, with pokkah boeng disease being one of the most damaging. Therefore, developing a sensitive method for the accurate detection of the pokkah boeng disease pathogen is crucial for ensuring the safety of sugar. This work presents a portable dual-modal detection device, assisted by a smartphone, which is based on MoS2@GDY, Mn3O4@Au nanomenzyme, cross-N DNA framework and Exo III exonuclease-assisted CHA signal amplification technology. The cross-N DNA framework provides many binding sites and is not restricted by AuNPs scattering positions, enhancing the signal output strength of the sensor. Additionally, the detection system incorporates a high-power-density capacitor to further amplify the electrochemical detection signal, increasing sensitivity by 9.1 times. Moreover, the use of electrochemical and colorimetric dual-mode detection effectively avoids mutual interference, reducing the likelihood of false positives from a single signal. Under optimized conditions, the proposed method has a linear range of 0.0001–10,000 pM, and with a detection limit of 6.1 aM (S/N=3). This high-sensitivity, high-reliability portable sensing method shows significant potential for the early detection and real-time on-site monitoring of the pokkah boeng disease pathogen.

The Srgb Color Space Based Density Analysis for Brain Tumor Segmentation

One of the important areas of study in medical image processing is early disease detection and diagnosis. A subfield of medical image processing includes the brain tumour segmentation procedure. Medical professionals have an efficient means of diagnosing illnesses because to computer vision and machine learning technology. The Srgb based density analysis is used in this study to isolate the area of brain tumours in MRI images. To differentiate the tumour area from the surrounding area, the intensity values of the input are normalised using the Gaussian filter and the Srgb colour space. In brain MRI samples, the adaptive threshold approach aids in locating potential tumour space. By calculating region parameters like area and density function, the actual space occupied by brain tumours is recovered. Ultimately, by applying morphological functions and removing potential false positives, the accurate tumour space is identified. Recall, accuracy, and F-measure are a few of the performance indicators that are used to evaluate how effective the suggested strategy.

Assessing CellaVision peripheral blood and advanced RBC application software's impact on hematologic morphology reporting: Real-world implementation experience in a large Veterans Affairs hospital.

This quality improvement study, conducted at the Kansas City Veterans Affairs Medical Center in Kansas City, Missouri, examined the change in patient reporting after transitioning from manual to CellaVision software-guided differentials and slide reviews. The primary focus was blasts and schistocytes, given their clinical significance. Three months of manual and CellaVision patient data were examined between May 2022 and February 2023. Blast and schistocyte patients were standardized to the total specimens performed and compared using 2-sample proportion testing. Other red blood cell (RBC) morphologies were also compared using this statistical analysis. Blast and schistocyte reporting after technologist review statistically increased after implementing the CellaVision software. This finding was most prevalent for low-percentage blasts. In addition, both morphologies experienced varying degrees of false-positive reporting, with 33% for low-percentage blasts and 91.7% for schistocytes. Other RBC morphologies displayed different levels of change, which could be clinically significant. The CellaVision software's increased sensitivity to blasts and schistocytes may benefit patient care, especially those with hematologic disorders. The software's high false-positive rate can be reduced by implementing quality metrics that prioritize clinically significant cells. This can be accomplished by implementing a CellaVision Champion to monitor reporting changes, perform patient lookbacks through the software, and provide technologist education. In addition, adopting a more stringent grading scale for schistocytes could also improve the high false-positive rate. Overall, CellaVision provides the ability to enhance hematology quality metrics by providing access to how patient cells are categorized and offering prompt education.

Alternative Perspectives on Obesity and Hypertension.

It is known that obesity and hypertension have a relationship with one another. Often, obesity is thought to directly cause hypertension, with a list of mechanisms commonly cited. This, however, does not do the relationship justice. Not only can the directionality of the relationship be flipped, but the mechanisms may be misattributed confounders, themselves. Beyond this, some argue that the results of trials using glucagon-like-peptide-1 receptor agonist (GLP1R) medications suggest a causal relationship between obesity and hypertension, but this will be debunked. The relationship is far from linear, and mainstream literature often excludes key confounders that will be discussed in this article including food insecurity, mental health, socioeconomic status (SES), and weight stigma and discrimination. The factors used to measure the risk of hypertension as well as the measurements of hypertension, itself, need to be reexamined. For instance, there may be a high amount of "false positives" among the diagnosed. Finally, current research needs to be critically evaluated for forms of weight centrism and weight bias, deciphering improper assumptions from true, evidence-based science.

Contribution of an Artificial Intelligence Tool in the Detection of Incidental Pulmonary Embolism on Oncology Assessment Scans

Introduction: The incidence of venous thromboembolism is estimated to be around 3% of cancer patients. However, a majority of incidental pulmonary embolism (iPE) can be overlooked by radiologists in asymptomatic patients, performing CT scans for disease surveillance, which may significantly impact the patient’s health and management. Routine imaging in oncology is usually reviewed with delayed hours after the acquisition of images. Nevertheless, the advent of AI in radiology could reduce the risk of the diagnostic delay of iPE by an optimal triage immediately at the acquisition console. This study aimed to determine the accuracy rate of an AI algorithm (CINA-iPE) in detecting iPE and the duration until the management of cancer patients in our center, in addition to describing the characteristics of patients with a confirmed pulmonary embolism (PE). Materials and Methods: This is a retrospective analysis of the role of Avicenna’s CE-certified and FDA-cleared CINA-iPE algorithm in oncology patients treated at Gustave Roussy Cancer Campus. The results obtained from the AI algorithm were compared with the attending radiologist’s report and were analyzed by both a radiology resident and a senior radiologist. In case of any discordant results, the reason for this discrepancy was further investigated. The duration between the exact time of the CT scan and analysis was assessed, as well as the duration from the result’s report and the start of active management. Results: Out of 3047 patients, 104 alerts were detected for iPE (prevalence of 1.3%), while 2942 had negative findings. In total, 36 of the 104 patients had confirmed PE, while 68 alerts were false positives. Only one patient reported as negative by the AI tool was deemed to have a PE by the radiologist. The sensitivity and specificity of the AI model were 97.3% and 97.74%, while the PPV and NPV were 34.62% and 99.97%, respectively. Most causes of FP were artifacts (22 cases, 32.3%) and lymph nodes (11 cases, 16.2%). Seven patients experienced delayed diagnosis, requiring them to return to the ER for treatment after being sent home following their scan. The remaining patients received prompt care immediately after their testing, with a mean delay time of 8.13 h. Conclusions: The addition of an AI system for the detection of unsuspected PEs on chest CT scans in routine oncology care demonstrated a promising efficacy in comparison to human performance. Despite a low prevalence, the sensitivity and specificity of the AI tool reached 97.3% and 97.7%, respectively, with detection of all the reported clinical PEs, except one single case. This study describes the potential synergy between AI and radiologists for an optimal diagnosis of iPE in routine clinical cancer care. Clinical relevance statement: In the oncology field, iPEs are common, with an increased risk of morbidity when missed with a delayed diagnosis. With the assistance of a reliable AI tool, the radiologist can focus on the challenging analysis of oncology results while dealing with urgent diagnosis such as PE by sending the patient straight to the ER (Emergency Room) for prompt treatment.

Evaluation of the Epitogen Lyme Detect IgG ELISA: a novel peptide multiplexing approach.

Lyme Borreliosis (LB), or Lyme disease, is a growing health concern caused by Borrelia burgdorferi sensu lato (Bbsl) bacteria transmitted through tick bites, and untreated cases can lead to severe health complications. Existing serology tests, while valuable, have low sensitivity in early infection stages where diagnosis is vital, interpretation variability, and false positives from cross-reactivity, while direct detection methods also suffer from low sensitivity, due to the inconsistent presence of Bbsl components in clinical samples. This study validated the diagnostic performance of the novel Epitogen Lyme Detect IgG enzyme-linked immunosorbent assay (ELISA) based on scaffold-displayed peptide antigens, using 120 specific immunodominant epitopes selected from 37 antigenic bacterial proteins corresponding to the main pathogenic Bbsl genospecies. Using 220 serum samples from Scottish patients with early, late, and disseminated LB, the assay's sensitivity was compared with that of the LIAISON Borrelia IgG CLIA, while specificity was assessed with 198 control samples, including healthy individuals and patients with diseases that are humorally similar. The Epitogen Lyme Detect IgG assay demonstrated comparable performance to the LIAISON Borrelia IgG in disseminated and late LB (Lyme neuroborreliosis, acrodermatitis chronica atrophicans, and Lyme arthritis). Notably, the Epitogen Lyme Detect IgG showed significantly higher sensitivity in patients with suspected erythema migrans, while maintaining high specificity. The Epitogen Lyme Detect IgG ELISA offers a promising advancement in LB diagnostics, demonstrating its potential for more accurate and timely diagnosis, particularly in the early stages of LB infection.IMPORTANCELyme Borreliosis (LB), caused by Borrelia burgdorferi sensu lato bacteria, poses significant health risks if undiagnosed or diagnosed late. Current diagnostic tests have limitations, especially in early-stage detection. This study validates the Epitogen Lyme Detect IgG enzyme-linked immunosorbent assay, demonstrating superior sensitivity in early LB detection while maintaining high specificity. The Epitogen Lyme Detect IgG comprises a suite of 120 immunodominant IgG epitopes/peptides from 37 bacterial antigens, covering the main LB-causing species: Borrelia burgdorferi sensu stricto, Borrelia afzelii, Borrelia garinii, and Borrelia mayonii. The novel design of multiplexing peptide antigens onto a scaffold to facilitate expression, correct folding, and orientation of the relevant peptides offers a promising advancement, potentially leading to more accurate and timely LB diagnoses and improving patient outcomes.

Simulation of 69 microbial communities indicates sequencing depth and false positives are major drivers of bias in prokaryotic metagenome-assembled genome recovery.

We hypothesize that sample species abundance, sequencing depth, and taxonomic relatedness influence the recovery of metagenome-assembled genomes (MAGs). To test this hypothesis, we assessed MAG recovery in three in silico microbial communities composed of 42 species with the same richness but different sample species abundance, sequencing depth, and taxonomic distribution profiles using three different pipelines for MAG recovery. The pipeline developed by Parks and colleagues (8K) generated the highest number of MAGs and the lowest number of true positives per community profile. The pipeline by Karst and colleagues (DT) showed the most accurate results (~ 92%), outperforming the 8K and Multi-Metagenome pipeline (MM) developed by Albertsen and collaborators. Sequencing depth influenced the accurate recovery of genomes when using the 8K and MM, even with contrasting patterns: the MM pipeline recovered more MAGs found in the original communities when employing sequencing depths up to 60 million reads, while the 8K recovered more true positives in communities sequenced above 60 million reads. DT showed the best species recovery from the same genus, even though close-related species have a low recovery rate in all pipelines. Our results highlight that more bins do not translate to the actual community composition and that sequencing depth plays a role in MAG recovery and increased community resolution. Even low MAG recovery error rates can significantly impact biological inferences. Our data indicates that the scientific community should curate their findings from MAG recovery, especially when asserting novel species or metabolic traits.

Augmenting Data Privacy Protocols and Enacting Regulatory Frameworks for Cryptocurrencies via Advanced Blockchain Methodologies and Artificial Intelligence

This study examines the effectiveness of current data privacy protocols within cryptocurrency platforms, focusing on encryption strength, anonymity techniques, and AI-powered regulatory compliance tools. Data were sourced from CoinMarketCap and Kaggle, including metrics like Bit Strength, Breach Incidents, and Anonymity Scores, which were analyzed using descriptive statistics, t-tests, and logistic regression. Results showed no significant relationship between encryption strength and breach incidents (p = 0.817), indicating that encryption strength may not be a primary factor in breach prevention. The weak correlation between encryption strength and breaches suggests that other elements, such as platform vulnerabilities or user behaviour, could play a more critical role in security. AI systems, evaluated through metrics like precision (0.168), recall (0.204), and F1 score (0.184), struggled with false positives, showing limitations in accurately detecting breaches and highlighting the need for more refined AI models. Advanced blockchain technologies like Zero-Knowledge Proofs and Homomorphic Encryption enhanced privacy but increased computational costs. It is recommended that hybrid encryption methods be adopted to balance privacy and performance and improve AI systems for more accurate breach detection. Governments must create clear regulations that encourage innovation while ensuring compliance.

Development and Evaluation of Real-Time Quantitative PCR Assays for Detection of Phellinus noxius Causing Brown Root Rot Disease.

Brown root rot disease (BRRD) is a highly destructive tree disease. Early diagnosis of BRRD has been challenging because the first symptoms and signs are often observed after extensive tissue colonization. Existing molecular detection methods, all based on the internal transcribed spacer (ITS) region, were developed without testing against global Phellinus noxius isolates, other wood-decay fungi, or host plant tissues. This study aimed to develop SYBR Green real-time quantitative PCR (qPCR) assays for P. noxius. The primer pair Pn_ITS_F/Pn_ITS_R targets the ITS, and the primer pair Pn_NLR_F/Pn_NLR_R targets a P. noxius-unique group of homologous genes identified through a comparative genomics analysis. The homologous genes belong to the nucleotide-binding-oligomerization-domain-like receptor (NLR) superfamily. The new primer pairs and a previous primer pair G1F/G1R were optimized for qPCR conditions and tested for specificity using 61 global P. noxius isolates, 5 other Phellinus species, and 22 non-Phellinus wood-decay fungal species. Although all three primer pairs could detect as little as 100 fg (approximately 2.99 copies) of P. noxius genomic DNA, G1F/G1R had the highest specificity and Pn_NLR_F/Pn_NLR_R had the highest efficiency. To avoid false positives, the cutoff quantification cycle values were determined as 34 for G1F/G1R, 29 for Pn_ITS_F/Pn_ITS_R, and 32 for Pn_NLR_F/Pn_NLR_R. We further validated these qPCR assays using Ficus benjamina seedlings artificially inoculated with P. noxius, six tree species naturally infected by P. noxius, rhizosphere soil, and bulk soil. The newly developed qPCR assays provide sensitive detection and quantification of P. noxius, which is useful for long-term monitoring of BRRD status.

Evaluating the Effectiveness of a Large Language Model in the Psychological Assessment of Potential Liver Transplant Candidates: A Feasibility Study

Introduction: The use of artificial intelligence (AI) in medical settings has shown promise in various domains including diagnostics, treatment recommendation, and patient management. Recent advances have explored the potential of AI in pre-surgical assessments, but its application in transplant psychology remains unexplored. The objective of this study is to assess the effectiveness of AI in the context of screening potential liver transplant donors and recipients. To assess the feasibility of using ChatGPT-4 to screen potential liver transplant donors and recipients. Methods: This study utilizes a cross-sectional research design to evaluate the feasibility of using ChatGPT-4 in the preliminary screening of living liver donors and liver transplant candidates. The study aims to determine the accuracy and reliability of ChatGPT-4 in assessing hypothetical scenarios involving potential donors and recipients. Results: The analysis showed no found no significant differences between ChatGPT-4 and the expert panel in assessing liver transplant candidates, demonstrating an overall accuracy of 83.58%, sensitivity of 56.10%, and specificity of 80.49%. Additionally, the Cohen’s Kappa statistic of 0.68 (95% CI: 0.52-0.83) indicated substantial agreement between ChatGPT-4 and the psychologists’ evaluations. The absence of false positives (0%) and a low false negative rate (8%) emphasize ChatGPT-4's cautious and accurate decision-making capabilities. Conclusion: The findings of this study demonstrate that ChatGPT-4 has the potential to serve as an effective screening tool for liver transplant candidates, complementing the work of human experts and enhancing the overall efficiency of the transplant process. While challenges remain, the integration of AI into the liver transplantation workflow could lead to significant improvements in candidate evaluation and patient outcomes, paving the way for the broader application of AI in clinical practice.

Mitigating Financial Fraud in Indonesia: A Grounded Theory Approach to Fraud Detection in P2P Lending, Rural Banks, and Conventional Banks

The rapid digitalization of Indonesia’s financial sector, driven by the growth of Peer-to-Peer Lending (P2PL) platforms, Rural Credit Banks (BPR), and conventional banks, has expanded financial inclusion but also heightened the risk of fraud. This study investigates the fraud detection systems (FDS) employed by these diverse financial institutions using a grounded theory approach. By conducting semi-structured interviews with key stakeholders—including directors, compliance managers, and fraud control officers—the research identifies critical factors influencing fraud detection in Indonesia’s financial ecosystem. The study reveals that while larger institutions leverage advanced machine learning models and real-time transaction monitoring, smaller entities like BPRs rely heavily on manual oversight due to resource constraints. Furthermore, the integration of multi-subject perception differences is highlighted as a novel and effective approach for fraud detection, particularly in P2PL platforms, by analyzing multiple data sources to reduce bias and increase accuracy. However, the study also uncovers significant challenges in implementing robust FDS, including high false-positive rates, regulatory complexities, and resource limitations, particularly in smaller institutions. The findings emphasize the need for a tailored fraud detection framework that addresses the unique needs of each sector, balancing technological advancements with operational feasibility.

Lightweight Advanced Deep Neural Network (DNN) Model for Early-Stage Lung Cancer Detection.

Background: Lung cancer, also known as lung carcinoma, has a high mortality rate; however, an early prediction helps to reduce the risk. In the current literature, various approaches have been developed for the prediction of lung carcinoma (at an early stage), but these still have various issues, such as low accuracy, high noise, low contrast, poor recognition rates, and a high false-positive rate, etc. Thus, in this research effort, we have proposed an advanced algorithm and combined two different types of deep neural networks to make it easier to spot lung melanoma in the early phases. Methods: We have used WDSI (weakly supervised dense instance-level lung segmentation) for laborious pixel-level annotations. In addition, we suggested an SS-CL (deep continuous learning-based deep neural network) that can be applied to the labeled and unlabeled data to improve efficiency. This work intends to evaluate potential lightweight, low-memory deep neural net (DNN) designs for image processing. Results: Our experimental results show that, by combining WDSI and LSO segmentation, we can achieve super-sensitive, specific, and accurate early detection of lung cancer. For experiments, we used the lung nodule (LUNA16) dataset, which consists of the patients' 3D CT scan images. We confirmed that our proposed model is lightweight because it uses less memory. We have compared them with state-of-the-art models named PSNR and SSIM. The efficiency is 32.8% and 0.97, respectively. The proposed lightweight deep neural network (DNN) model archives a high accuracy of 98.2% and also removes noise more effectively. Conclusions: Our proposed approach has a lot of potential to help medical image analysis to help improve the accuracy of test results, and it may also prove helpful in saving patients' lives.

Diagnostic Utility of the "Air Sign" as a Radiological Indicator for Mandibular Body and Angle Fractures.

Background: To plan optimal treatment, obtain satisfactory outcomes, and avoid undesirable clinical errors, surgeons need to have efficient tools for providing a complete and prompt diagnosis. The aim of this study was to establish the sensitivity, specificity, positive and negative predictive values, false positive rate, and false negative rate of the "air sign" (AS) within soft tissues as an indirect radiological indicator of mandibular body and angle fractures. Methods: A retrospective analysis of preoperative computed tomography (CT) and cone beam computed tomography (CBCT) scans was performed on patients with mandibular fractures within a three-year period. Two fracture types were analyzed: open and closed fractures. Results: Forty-three patients with a total of 71 mandibular fractures were included in the study. The mean age of the patients was 35 years, and the majority were male (83.7%). The sensitivity of the AS was 92.2%, specificity = 90.0%, positive predictive value = 95.9%, negative predictive value = 81.8%, false positive rate = 10.0%, and false negative rate = 7.8%. Higher values were observed for open fractures compared to closed fractures. Conclusions: The sensitivity and specificity of AS are lower than those of OPG, CT, and CBCT. However, AS offers an important additional radiological indicator that can effectively reduce the risk of misdiagnosing mandibular body and angle fractures.

PONYTA: prioritization of phenotype-related genes from mouse KO events using PU learning on a biological network.

Transcriptome data from gene knock-out (KO) experiments in mice provide crucial insights into the intricate interactions between genotype and phenotype. Differentially expressed gene (DEG) analysis and network propagation (NP) are well-established methods for analyzing transcriptome data. To determine genes related to phenotype changes from a KO experiment, we need to choose a cutoff value for the corresponding criterion based on the specific method. Using a rigorous cutoff value for DEG analysis and NP is likely to select mostly positive genes related to the phenotype, but many will be rejected as false negatives. On the other hand, using a loose cutoff value for either method is prone to include a number of genes that are not phenotype-related, which are false positives. Thus, the research problem at hand is how to deal with the trade-off between false negatives and false positives. We propose a novel framework called PONYTA for gene prioritization via positive-unlabeled (PU) learning on biological networks. Beginning with the selection of true phenotype-related genes using a rigorous cutoff value for DEG analysis and NP, we address the issue of handling false negatives by rescuing them through PU learning. Evaluations on transcriptome data from multiple studies show that our approach has superior gene prioritization ability compared to benchmark models. Therefore, PONYTA effectively prioritizes genes related to phenotypes derived from gene KO events and guides in vitro and in vivo gene KO experiments for increased efficiency. The source code of PONYTA is available at https://github.com/Jun-Hyeong-Kim/PONYTA.

YOLO-GE: An Attention Fusion Enhanced Underwater Object Detection Algorithm

Underwater object detection is a challenging task with profound implications for fields such as aquaculture, marine ecological protection, and maritime rescue operations. The presence of numerous small aquatic organisms in the underwater environment often leads to issues of missed detections and false positives. Additionally, factors such as the water quality result in weak target features, which adversely affect the extraction of target feature information. Furthermore, the lack of illumination underwater causes image blur and low contrast, thereby increasing the difficulty of the detection task. To address these issues, we propose a novel underwater object detection algorithm called YOLO-GE (GCNet-EMA). First, we introduce an image enhancement module to mitigate the impact of underwater image quality issues on the detection task. Second, a high-resolution feature layer is added into the network to improve the problems of missed detections and false positives for small targets. Third, we propose GEBlock, an attention-based fusion module that captures long-range contextual information and suppresses noise from lower-level feature layers. Finally, we combine an adaptive spatial fusion module with the detection head to filter out conflicting feature information from different feature layers. Experiments on the UTDAC2020, DUO and RUOD datasets show that the proposed method achieves an optimal detection accuracy.

International external quality assessment study for detection of monkeypox virus by PCR supporting laboratory preparedness during the 2022–2023 mpox outbreak and beyond

BackgroundDiagnostic capabilities and correspondent External Quality Assessments (EQA) are key for outbreak preparedness. To support diagnostic facilities with a quality assessment of newly established monkeypox virus (MPXV) molecular diagnostic workflows, Quality Control for Molecular Diagnostics (QCMD) and the Bundeswehr Institute of Microbiology (IMB) piloted an international EQA study conducting four challenges from autumn 2022 to summer 2023 during the global mpox outbreak. ObjectivesTo assess the performance (sensitivity/specificity) of molecular assays used by diagnostic laboratories. Study designInactivated EQA panels were prepared and distributed containing seven samples of clade Ia and clade IIb MPXV strains at different viral concentrations, two specificity controls with other zoonotic orthopoxviruses (vaccinia and cowpox virus) and a negative control. Assessment was based on reported qualitative testing results. ResultsIn this outbreak-related EQA study, a total of 192 laboratories from 37 countries reported 346 qualitative datasets. Overall, core samples were correctly detected by approximately 92 % of participants in all four challenges. While sensitivity performance was acceptable in at least 90 % of datasets correctly reported even for educational MPXV-positive samples with low viral concentration [102 genome equivalents (GE)/mL], several laboratories reported the educational specificity controls as false positives or were unable to differentiate MPXV from related orthopoxviruses. ConclusionsMpox is now a globally occurring infection with a demand for quality-assured diagnostic capabilities. The newly established EQA scheme presented here, offers a multi-purpose panel for orthopoxviruses with a focus on MPXV which will continue to ensure diagnostic quality in clinical settings with up-to-date sample panels.

Integrated Genetic Algorithm and Deep Learning Approach for Effective Cyber-Attack Detection and Classification in Industrial Internet of Things (IIoT) Environments

AbstractCyber-attack detection within Industrial Internet of Things (IIoT) environments presents unique challenges due to the complex, resource-constrained, and real-time nature of these networks. Traditional detection techniques often struggle to adapt to the dynamic environment of IIoT. For instance, many existing methods rely on signature-based detection, which fails to identify evolving threats. Other approaches, such as anomaly-based detection, can generate a high rate of False Positives, leading to inefficiencies in threat management. To address these challenges, we propose a novel detection and classification model specifically tailored for IIoT environments. The proposed model integrates Genetic Algorithms (GA) and Deep Learning (DL) to enhance cyber-attack detection within IIoT environments. The GA component optimises feature selection from raw network data, ensuring the extraction of meaningful and relevant features. Leveraging these selected features, the DL component constructs a robust model capable of accurately detecting and classifying various cyber-attack patterns across IIoT devices. Through experimentation on real-world IIoT network traffic (UNSW-NB 15 dataset), the proposed approach demonstrates its efficacy in improving attack detection accuracy and adaptability. The integration of GA and DL offers a synergistic solution that addresses the complexities of IIoT cybersecurity, contributing to a more secure and resilient IIoT ecosystem. The integrated GA–DL classification model developed in this work achieved 98% precision, 96% accuracy, 94% recall, and 12% losses with only less than 50% of the features of the UNSW-NB 15 dataset. The reduction in features required for the identification and classification of cyber-attacks reduces the processing time by 50%.

Explainable AI-based innovative hybrid ensemble model for intrusion detection

AbstractCybersecurity threats have become more worldly, demanding advanced detection mechanisms with the exponential growth in digital data and network services. Intrusion Detection Systems (IDSs) are crucial in identifying illegitimate access or anomalous behaviour within computer network systems, consequently opposing sensitive information. Traditional IDS approaches often struggle with high false positive rates and the ability to adapt embryonic attack patterns. This work asserts a novel Hybrid Adaptive Ensemble for Intrusion Detection (HAEnID), an innovative and powerful method to enhance intrusion detection, different from the conventional techniques. HAEnID is composed of a string of multi-layered ensemble, which consists of a Stacking Ensemble (SEM), a Bayesian Model Averaging (BMA), and a Conditional Ensemble method (CEM). HAEnID combines the best of these three ensemble techniques for ultimate success in detection with a considerable cut in false alarms. A key feature of HAEnID is an adaptive mechanism that allows ensemble components to change over time as network traffic patterns vary and new threats appear. This way, HAEnID would provide adequate protection as attack vectors change. Furthermore, the model would become more interpretable and explainable using Shapley Additive Explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME). The proposed Ensemble model for intrusion detection on CIC-IDS 2017 achieves excellent accuracy (97-98%), demonstrating effectiveness and consistency across various configurations. Feature selection further enhances performance, with BMA-M (20) reaching 98.79% accuracy. These results highlight the potential of the ensemble model for accurate and reliable intrusion detection and, hence, is a state-of-the-art choice for accuracy and explainability.

Diagnostic performance of Leishmania braziliensis and Leishmania peruviana antigens in the immunoblot method for the detection of american tegumentary leishmaniasis.

Motivation for the study. To contribute to the immunogenic character of soluble and excretion/secretion antigens of Leishmania braziliensis and Leishmania peruviana with the aim of identifying proteins with diagnostic potential. Main findings. The soluble antigen of Leishmania braziliensis has a sensitivity in the detection of ATL of 87.7%, specificity of 100% and a false positive rate of 20% against sera from patients with Chagas disease and 8.3% with mycosis. Implications. Immunoblot can improve the resolution capacity in the serological diagnosis of American tegumentary Leishmaniasis, particularly in patients where the length of the disease and the clinical form make difficult the diagnosis by parasitological methods. This study aimed to determine the performance of Leishmania braziliensis and Leishmania peruviana antigens in the detection of ATL by using serum samples obtained between 2013 - 2016. The obtained soluble and excretion/secretion antigens were transferred to membrane nitrocellulose by immunoblot assay. The evaluation was carried out against sera confirmed for ATL, at a confidence level of 95%, determining that the soluble antigen of Leishmania braziliensis had a sensitivity of 87.7%, specificity of 100% and area under the curve of 0.95; on the other hand, Leishmania peruviana showed values of 92.3%, 95.7% and 0.94, respectively. According to the results, we recommend that the reported immunogenic regions should be characterized and analyzed in order to continue with the development of recombinant and synthetic proteins, aimed at improving the efficiency of the serological diagnosis of the disease.

State monitoring and fault prediction of centrifugal compressors based on long short–term memory and principal component analysis (LSTM-PCA)

Centrifugal compressors are widely used in the petroleum and natural gas industry for gas compression, reinjection, and transportation. Early fault identification and fault evolution prediction for centrifugal compressors can improve equipment safety and reduce maintenance and operating costs. This article proposes a dynamic process monitoring method for centrifugal compressors based on long short-term memory (LSTM) and principal component analysis (PCA). This method constructs a sliding window for monitoring at each sampling point, which contains 100 data from the past and current time points, and uses LSTM to predict 30 future data points. At the same time, this method is also combined with the PCA threshold process monitoring method to construct a new LSTM-PCA monitoring algorithm. And the method was validated using centrifugal compressor process data. The results show that this method can effectively detect process anomalies, The improvements significantly reduced the false positive rate of detected anomalies, and can make multi-step advance predictions of system behavior after faults occur.