Effective Detection Method Research Articles

Supervised contrastive pre-training models for mammography screening

Breast cancer is now the most deadly cancer worldwide. Mammography screening is the most effective method for early detection and diagnosis of breast cancer. Due to the lack of labeled mammograms, building an AI system for mammography screening often relies heavily on human-designed data augmentation, which doesn’t always perform robustly when applied to clinical scenarios. This paper presents a novel framework of Supervised Contrastive Pre-training followed by Supervised Fine-tuning (SCP+SF) for mammography screening. Unlike the previous approaches, the proposed supervised contrastive pre-training does not need a data augmentation module. We apply the SCP+SF framework to two challenging and important mammography screening tasks for breast cancer: mammographic abnormality screening and mammographic malignancy screening. Our extensive experiments on a large-scale dataset show that the supervised contrastive pre-training (SCP) can substantially improve the final model performance compared with the traditional direct supervised training approach. Superior results of AUC and specificity/sensitivity have been achieved on two clinically significant mammographic screening tasks in comparison with previously reported State-Of-The-Art approaches. We believe this work is the first to show that supervised contrastive pre-training (SCP) followed by supervised fine-tuning (SF) can outperform the supervised counterpart on these two critical medical imaging tasks.

A miniaturized RPA-CRISPR/Cas12a-based nucleic acid diagnostic platform for rapid and simple self-testing of SARS-CoV-2.

Nucleic acid testing is the most effective detection method currently available for the diagnosis of respiratory infectious diseases. However, the conventional real-time fluorescent quantitative PCR technique, which is regarded as the gold standard method for nucleic acid detection, presents significant challenges for implementation in home self-testing and popularization in underdeveloped regions due to its rigorous experimental standards. It is therefore clear that an easy-to-use, miniaturized nucleic acid testing technology and products for nonprofessionals are of great necessity to define the pathogens and assist in controlling disease transmission. (87) RESULTS: In this study, we propose a strategy for self-testing of respiratory pathogen nucleic acid that is oriented towards the public and user-friendly. The proposed system integrates the processes of extraction-free nucleic acid release, RPA isothermal amplification, and CRISPR fluorescence detection into a compact configuration. A microfluidic testing chip actuated by air pouches and a battery/USB-powered reusable device has been developed to enable simultaneous detection of internal reference genes and viral targets in a fully enclosed condition. The system allows for sample-in, and result-out testing in less than 30min with a detection limit of 2 copies/μL. Additionally, a straightforward signal-light-based result display method has been developed to make it easy and intuitive for users to access the results. Furthermore, freeze-drying reagent is introduced to guarantee the storage and transportation of testing chips in ambient conditions. (135) SIGNIFICANCE: This work presents a miniaturized, portable, and highly sensitive nucleic acid detection system, where simple operating procedures have been designed for unskilled users. It is our belief that the testing system developed in this work is well suited for home-based self-testing and infection diagnosis in resource-limited areas, due to the above-mentioned advantages. (52).

Detection of Staphylococcus aureus via IgY-Based immunomagnetic separation and a phage lysin LysGH15-Based colloidal gold immunochromatographic assay.

Staphylococcus aureus (S. aureus) is an important human pathogen that frequently causes disease in hospital and community settings. Due to the pervasive emergence of virulent and multidrug-resistant methicillin-resistant S. aureus (MRSA) strains, S. aureus infections have become a leading cause of morbidity and mortality. We aimed to develop a reliable and effective detection method that can be used for preventing and managing infections caused by S. aureus. A novel S. aureus detection platform that combines immunomagnetic separation via immunomagnetic beads (IMBs) and the phage lysin LysGH15-based colloidal gold immunochromatographic assay (GICA) (named IMBs-GICA) was developed. IMBs-GICA was performed on 426 consecutive samples, including 185, 107, 74 and 60 samples from sputum, urine, bronchoalveolar lavage fluid (BALF) and serum samples, respectively. S. aureus culture and PCR assays were routinely conducted in parallel. By targetingS. aureus, the IMBs-GICA platform could detect as little as 40 CFU/mL within 35 min. Among the clinical specimens, 38 samples were S. aureus culture positive, and 36 were IMBs-GICA positive, resulting in an IMBs-GICA sensitivity of 89.5 % (95 % confidence interval 74.3-96.6) and a specificity of 99.5 % (97.9-99.9). IMBs-GICA can rapidly and sensitively detect for S. aureus, demonstrating the potential utility of this developed method in the field of testing.

A Survey on Liver Cancer Detection Using Hyperfusion of CNN and SVM in Machine Learning

Since liver cancer ranks among of the most aggressive renditions of the disease, improving patient outcomes requires early identification. We propose an inventive tactic to liver cancer detection by integrating CNN and SVM. CNNs, known for their powerful feature extraction capabilities, are particularly effective in analysing complex medical images. SVMs, on the other hand, are efficient classifiers that can separate data points in high-dimensional spaces with accuracy. By merging the feature extraction strength of CNN with the classification efficiency of SVM, the proposed model aims to enhance liver cancer detection accuracy and robustness. The experimental results reveal that the fused CNN-SVM model significantly surpasses the performance of standalone CNN and SVM models, achieving a high detection accuracy of 95.2%. This hybrid method offers a promising direction for improving the precision of computer-aided diagnosis systems, contributing to more effective and reliable liver cancer detection methods that can assist healthcare professionals in making timely decisions.

Long Non-Coding RNAs in Ovarian Cancer: Mechanistic Insights and Clinical Applications

Background/Objectives: Ovarian cancer is a leading cause of gynecological cancer mortality worldwide, often diagnosed at advanced stages due to vague symptoms and the lack of effective early detection methods. Long non-coding RNAs (lncRNAs) have emerged as key regulators in cancer biology, influencing cellular processes such as proliferation, apoptosis, and chemoresistance. This review explores the multifaceted roles of lncRNAs in ovarian cancer pathogenesis and their potential as biomarkers and therapeutic targets. Methods: A comprehensive literature review was conducted to analyze the structural and functional characteristics of lncRNAs and their contributions to ovarian cancer biology. This includes their regulatory mechanisms, interactions with signaling pathways, and implications for therapeutic resistance. Advanced bioinformatics and omics approaches were also evaluated for their potential in lncRNA research. Results: The review highlights the dual role of lncRNAs as oncogenes and tumor suppressors, modulating processes such as cell proliferation, invasion, and angiogenesis. Specific lncRNAs, such as HOTAIR and GAS5, demonstrate significant potential as diagnostic biomarkers and therapeutic targets. Emerging technologies, such as single-cell sequencing, provide valuable insights into the tumor microenvironment and the heterogeneity of lncRNA expression. Conclusions: LncRNAs hold transformative potential in advancing ovarian cancer diagnosis, prognosis, and treatment. Targeting lncRNAs or their associated pathways offers promising strategies to overcome therapy resistance and enhance personalized medicine. Continued research integrating omics and bioinformatics will be essential to unlock the full clinical potential of lncRNAs in ovarian cancer management.

Enhanced Deep Stacked CapsNet Ensemble Gazelle Neural Network for multi‐level fabric defect classification

AbstractFabric detection in the materials industry plays a vital role in the global economy, making effective quality control measures essential to ensure product value and reduce manufacturing waste. With the rise of Industry 4.0, manufacturing companies have been striving to develop automated fabric defect detection systems to overcome the limitations of traditional manual inspection. However, because of challenges in creating highly effective fabric defect detection methods with strong noise resistance, conventional systems often struggle to capture intricate fabric details and accurately distinguish between defect types. To address these challenges, this research introduces an innovative approach called the Enhanced Deep Stacked capsNet Ensemble Gazelle Neural Network (EDSEGNN) for multi‐level local and global defect classification. By incorporating the Window‐aware Guided Image Filtering technique, image quality and resolution are enhanced, enabling the detection of fine fabric details. Additionally, the Wavelet Packet Transform aids in segmenting fabric defects by identifying small patterns through varying frequency waves. In the final stage, the EDSEGNN model performs local defect identification and multi‐level global defect classification, distinguishing between normal and defective patterns while categorising defect types like discoloration, stains, foreign objects, cuts, holes, thread issues and metal contamination. The proposed method achieves impressive results, with a peak accuracy of 99.8%, along with high recall (99.5%) and F1‐score (99%), compared with existing methods. The proposed approach offers a highly accurate and robust solution to the challenges faced by traditional fabric defect detection systems, representing a valuable advance in automated quality control for the materials industry.

Detecting Malicious .NET Executables Using Extracted Methods Names

The .NET framework is widely used for software development, making it a target for a significant number of malware attacks by developing malicious executables. Previous studies on malware detection often relied on developing generic detection methods for Windows malware that were not tailored to the unique characteristics of .NET executables. As a result, there remains a significant knowledge gap regarding the development of effective detection methods tailored to .NET malware. This work introduces a novel framework for detecting malicious .NET executables using statically extracted method names. To address the lack of datasets focused exclusively on .NET malware, a new dataset consisting of both malicious and benign .NET executable features was created. Our approach involves decompiling .NET executables, parsing the resulting code, and extracting standard .NET method names. Subsequently, feature selection techniques were applied to filter out less relevant method names. The performance of six machine learning models—XGBoost, random forest, K-nearest neighbor (KNN), support vector machine (SVM), logistic regression, and naïve Bayes—was compared. The results indicate that XGBoost outperforms the other models, achieving an accuracy of 96.16% and an F1-score of 96.15%. The experimental results show that standard .NET method names are reliable features for detecting .NET malware.

Taming vision transformers for clinical laryngoscopy assessment.

Laryngoscopy, essential for diagnosing laryngeal cancer (LCA), faces challenges due to high inter-observer variability and the reliance on endoscopist expertise. Distinguishing precancerous from early-stage cancerous lesions is particularly challenging, even for experienced practitioners, given their similar appearances. This study aims to enhance laryngoscopic image analysis to improve early screening/detection of cancer or precancerous conditions. We propose MedFormer, a laryngeal cancer classification method based on the Vision Transformer (ViT). To address data scarcity, MedFormer employs a customized transfer learning approach that leverages the representational power of pre-trained transformers. This method enables robust out-of-domain generalization by fine-tuning a minimal set of additional parameters. MedFormer exhibits sensitivity-specificity values of 98%-89% for identifying precancerous lesions (leukoplakia) and 89%-97% for detecting cancer, surpassing CNN counterparts significantly. Additionally, when compared to the two selected ViT-based models, MedFormer also demonstrates superior performance. It also outperforms physician visual evaluations (PVE) in certain scenarios and matches PVE performance in all cases. Visualizations using class activation maps (CAM) and deformable patches demonstrate MedFormer's interpretability, aiding clinicians in understanding the model's predictions. We highlight the potential of visual transformers in clinical laryngoscopic assessments, presenting MedFormer as an effective method for the early detection of laryngeal cancer.

Out-of-Distribution Detection via outlier exposure in federated learning.

Among various out-of-distribution (OOD) detection methods in neural networks, outlier exposure (OE) using auxiliary data has shown to achieve practical performance. However, existing OE methods are typically assumed to run in a centralized manner, and thus are not feasible for a standard federated learning (FL) setting where each client has low computing power and cannot collect a variety of auxiliary samples. To address this issue, we propose a practical yet realistic OE scenario in FL where only the central server has a large amount of outlier data and a relatively small amount of in-distribution (ID) data is given to each client. For this scenario, we introduce an effective OE-based OOD detection method, called internal separation & backstage collaboration, which makes the best use of many auxiliary outlier samples without sacrificing the ultimate goal of FL, that is, privacy preservation as well as collaborative training performance. The most challenging part is how to make the same effect in our scenario as in joint centralized training with outliers and ID samples. Our main strategy (internal separation) is to jointly train the feature vectors of an internal layer with outliers in the back layers of the global model, while ensuring privacy preservation. We also suggest an collaborative approach (backstage collaboration) where multiple back layers are trained together to detect OOD samples. Our extensive experiments demonstrate that our method shows remarkable detection performance, compared to baseline approaches in the proposed OE scenario.

Biotinylated Viscosity Sensitive Cell Membrane Probe for Targeted Imaging and Precise Visualization of Tumor Cells and Tumors.

Cancer is a global health challenge that urgently requires more sensitive and effective cancer detection methods. Fluorescence imaging with small molecule fluorescent probes has shown great promise for cancer detection but most of the developed probes lack active tumor cell targeting, which makes them unable to selectively target tumors, thereby reducing the accuracy of in vivo tumor detection. Herein, we report a novel probe Bio-S that combines a viscosity-sensitive and cell membrane targetable fluorescent group with biotin for targeted imaging and precise visualization of tumor cells and tumors. Bio-S exhibits sensitive fluorescence changes for viscosity at ∼660 nm and excellent cell membrane localization and imaging ability (red fluorescence, wash-free, and long-term imaging). Moreover, compared with the nonbiotinylated control probe C6-S, the biotinylated Bio-S can specifically target tumor cell membranes, thereby achieving much higher selectivity and sensitivity in distinguishing tumor cells from normal cells. Mice imaging experiments show that tail vein injection of Bio-S can target tumors and monitor lung cancer metastasis at the in vivo level. Therefore, this work provides an effective new strategy and tool for tumor-targeted detection and precise diagnosis.

Research on the problem of landmines and explosive remnants of war worldwide and methods of demining

The study addresses the global issue of landmines, with a particular focus on the challenges in Bosnia and Herzegovina. The research is grounded in precise data and facts from institutions and accredited demining organizations. The study further concentrates on innovative and highly effective detection methods, specifically for areas with challenging topography like Bosnia and Herzegovina, concluding with an innovative application of GPR and LiDAR radar technologies. Additionally, it explores the most effective method for mine removal, and mechanical demining, providing a detailed overview of the operation and characteristics associated with this approach. Conclusively, the study proposes the most efficient solution: a combined system of flails and tillers that alternately complement each other to achieve complete mine clearance.

Flexible Site-Specific Labeling-Mediated Self-Assembly Sensor Based on Quantum Dots and LUMinescent AntiBody Sensor for Duplexed Detection of Antibodies.

Over recent years, the LUMinescent AntiBody Sensor (LUMABS) system, utilizing bioluminescence resonance energy transfer (BRET), has emerged as a highly effective method for antibody detection. This system incorporates NanoLuc (Nluc) as the donor and fluorescent protein (FP) as the acceptor. However, the limited Stokes shift of FP poses a challenge, as it leads to significant spectral cross-talk between the excitation and emission spectra. This issue complicates the implementation of multiplexed detection. To address this challenge, we present an innovative enhancement to the LUMABS sensor with quantum dots (QDs) as the acceptor instead of FP. The use of QDs offers several advantages over those of traditional FP-based sensors. The biotin-avidin system facilitates the flexible interchangeability of QDs, allowing for a more convenient multicolor sensor construct. The new QD-LUMABS system overcomes the limitations of spectral cross-talk and provides better spectral separation. This breakthrough enables the successful implementation of multiplexed detection for multiple targets simultaneously. Results demonstrated that the wavelength-tunable QD-LUMABS sensors achieved picomolar-level detection limits for antibodies and that this sensor-construction strategy was generally applicable among various epitopes and their antibodies. Furthermore, this sensor displayed excellent duplexing capabilities. These features underscore its potential for future clinical disease diagnosis applications.

PC-YOLO11s: A Lightweight and Effective Feature Extraction Method for Small Target Image Detection.

Compared with conventional targets, small objects often face challenges such as smaller size, lower resolution, weaker contrast, and more background interference, making their detection more difficult. To address this issue, this paper proposes an improved small object detection method based on the YOLO11 model-PC-YOLO11s. The core innovation of PC-YOLO11s lies in the optimization of the detection network structure, which includes the following aspects: Firstly, PC-YOLO11s has adjusted the hierarchical structure of the detection network and added a P2 layer specifically for small object detection. By extracting the feature information of small objects in the high-resolution stage of the image, the P2 layer helps the network better capture small objects. At the same time, in order to reduce unnecessary calculations and lower the complexity of the model, we removed the P5 layer. In addition, we have introduced the coordinate spatial attention mechanism, which can help the network more accurately obtain the spatial and positional features required for small targets, thereby further improving detection accuracy. In the VisDrone2019 datasets, experimental results show that PC-YOLO11s outperforms other existing YOLO-series models in overall performance. Compared with the baseline YOLO11s model, PC-YOLO11s mAP@0.5 increased from 39.5% to 43.8%, mAP@0.5:0.95 increased from 23.6% to 26.3%, and the parameter count decreased from 9.416M to 7.103M. Not only that, we also applied PC-YOLO11s to tea bud datasets, and experiments showed that its performance is superior to other YOLO-series models. Experiments have shown that PC-YOLO11s exhibits excellent performance in small object detection tasks, with strong accuracy improvement and good generalization ability, which can meet the needs of small object detection in practical applications.

Smart Accident Detection using IoT Technology

Road accidents and emergency services delay are the main significant issues. To overcome these issues need to develop a system. Efficient handling of accidents through the immediate detection and provide timely aid are more crucial. Accident detection and emergency system depends on IoT (Internet of things) with minimum delay are gaining significant attention towards industry and academic literature. Several researches are investigated using IOT technology to detect accidents. In this work, we proposed an effective accident detection method by employing five sensors not only to detect accident but also to report type of accident such as collision, no accident, roll over or fall off. In addition to that, the status of the accident is communicated to the IBM Watson Cloud platform. The incoming data received in the node red platform is integrated with the Google Maps to show location and other information about the accident that can be accessed by the hospital through website and also sending alert messages to victim acquaintances. In addition, two Machine Learning (ML) models based on K-Nearest Neighbor (KNN) model and the Naïve Bayes (NB) model are compared to find out the best accident detection model. It is noticed that the KNN model is the very effective ML model, which employed to know the accident status and also to enhance the system by providing patient’s details, a kill switch and sending messages often till acknowledgement is received.

How Good (Or Bad) Are LLMs at Detecting Misleading Visualizations?

In this study, we address the growing issue of misleading charts, a prevalent problem that undermines the integrity of information dissemination. Misleading charts can distort the viewer's perception of data, leading to misinterpretations and decisions based on false information. The development of effective automatic detection methods for misleading charts is an urgent field of research. The recent advancement of multimodal Large Language Models (LLMs) has introduced a promising direction for addressing this challenge. We explored the capabilities of these models in analyzing complex charts and assessing the impact of different prompting strategies on the models' analyses. We utilized a dataset of misleading charts collected from the internet by prior research and crafted nine distinct prompts, ranging from simple to complex, to test the ability of four different multimodal LLMs in detecting over 21 different chart issues. Through three experiments-from initial exploration to detailed analysis-we progressively gained insights into how to effectively prompt LLMs to identify misleading charts and developed strategies to address the scalability challenges encountered as we expanded our detection range from the initial five issues to 21 issues in the final experiment. Our findings reveal that multimodal LLMs possess a strong capability for chart comprehension and critical thinking in data interpretation. There is significant potential in employing multimodal LLMs to counter misleading information by supporting critical thinking and enhancing visualization literacy. This study demonstrates the applicability of LLMs in addressing the pressing concern of misleading charts.

Ultrasensitive detection of E. coli using bioinspired based platform.

Bacterial infections are a leading cause of mortality worldwide, underscoring the urgent need for effective detection methods. This study introduces a novel approach that combines surface-enhanced Raman spectroscopy (SERS) with an electro-optic technique for bacterial detection. The method utilizes a metal-semiconductor substrate that, when activated by an external electric field, significantly amplifies the SERS signal intensity. We validated this approach through a proof-of-concept study, demonstrating that the SERS signal of Gram-negative Escherichia coli can be enhanced tenfold by applying an electric field, confirming the method's efficacy for bacterial detection. Our findings highlight the potential of this rapid, label-free biosensor for pathogen detection with near single-cell sensitivity.

Development of a nanobody-based immunoassay for the detection of Escherichia coli O157:H7 in food samples.

Rapid and accurate detection of foodborne pathogens is crucial for food safety. E. coli O157:H7, a major foodborne pathogen, can cause severe illness even at low concentration, highlighting the necessity for effective detection methods. This study focuses on developing nanobodies specific to E. coli O157:H7, aiming to provide a more efficient, cost-effective, and stable analytical reagents which can be available internationally. An alpaca was immunized and a phage-displayed nanobody library was constructed. Following panning, nanobodies were expressed and characterized for their affinity and specificity. A nanobody-based immunoassay was developed, with a limit of detection of 8.7×103cfu/mL. The matrix effect was evaluated, and the assay successfully detected E. coli O157:H7 in various food samples (with 5cfu spiked), including ground beef, orange juice, and milk. The study developed a nanobody-based immunoassay for E. coli O157:H7 detection, underscoring the practical applicability of nanobodies for testing food safety monitoring.

Clustering- and statistic-based approach for detection and impact evaluation of faults in end-user substations of thermal energy systems

In response to climate change mitigation efforts, improving the efficiency of heat networks is becoming increasingly important. An efficient operation of energy systems depends on faultless performance. Following the need for effective fault detection and elimination methods, this study suggests a three-step workflow for increasing automation in managing defective substations on the user level within heat networks. The work focuses on a model region in northern Germany. The local heat network provides data in roughly hourly intervals, including the supply and return temperatures and the volume flow of the substations. Firstly, this study identifies common indicators of faults using k-means clustering analysis of the temperature data and expert knowledge: an exceeded return temperature level, very low cooling, and inverted temperature readings. With these indicators, the subsequent statistical identification approach confirms the successful detection of affected substations, with common diagnoses including disabled return temperature limitation units, defective motoric valves, and faults in the storage control. Lastly, the study evaluates the impact of faults on the system efficiency. Combining the temperature and the volume flow data, the workflow quantifies the negative influence of a fault, enabling the prioritization of fault elimination measures in practical application to enhance the overall system efficiency.

Performance Analysis Using CNN for Detecting Wood Knots

This study proposes a Convolutional Neural Network (CNN) model to quickly and accurately detect wood deformations. The performance of the CNN was enhanced by extracting structural deformation features, optimizing training parameters, and improving datasets. Experimental analyses demonstrate that the CNN achieved high accuracy rates and is an effective method for deformation detection. The CNN model was designed to identify various wood deformations. Its layered architecture was optimized to analyze deformations at different scales and levels of detail. Minimal preprocessing was applied to the images used during training, and data augmentation techniques were employed to enhance dataset diversity. The model was trained on a training dataset and tested on a validation dataset. Metrics such as loss function and accuracy were monitored throughout the training process. The CNN achieved an accuracy rate of 99.90% on the training dataset. This study highlights that the CNN model is an effective method for non-destructive detection of wood deformations. The proposed CNN model has potential applications in wood deformation detection and quality control processes.

Comparative evaluation of spatiotemporal methods for effective dengue cluster detection with a case study of national surveillance data in Thailand

Dengue fever poses a significant public health burden in tropical regions, including Thailand, where periodic epidemics strain healthcare resources. Effective disease surveillance is essential for timely intervention and resource allocation. Various methods exist for spatiotemporal cluster detection, but their comparative performance remains unclear. This study compared spatiotemporal cluster detection methods using simulated and real dengue surveillance data from Thailand. A simulation study explored diverse disease scenarios, characterized by varying magnitudes and spatial-temporal patterns, while real data analysis utilized monthly national dengue surveillance data from 2018 to 2020. Evaluation metrics included accuracy, sensitivity, specificity, positive predictive value, and negative predictive value. Bayesian models and FlexScan emerged as top performers, demonstrating superior accuracy and sensitivity. Traditional methods such as Getis Ord and Moran’s I showed poorer performance, while other scanning-based approaches like spatial SaTScan exhibited limitations in positive predictive value and tended to identify large clusters due to the inflexibility of its scanning window shape. Bayesian modeling with a space–time interaction term outperformed testing-based cluster detection methods, emphasizing the importance of incorporating spatiotemporal components. Our study highlights the superior performance of Bayesian models and FlexScan in spatiotemporal cluster detection for dengue surveillance. These findings offer valuable guidance for policymakers and public health authorities in refining disease surveillance strategies and resource allocation. Moreover, the insights gained from this research could be valuable for other diseases sharing similar characteristics and settings, broadening the applicability of our findings beyond dengue surveillance.