Binary Classification Research Articles

On most informative regions for binary classification of schizophrenia based on resting state fMRI data done by selection of functionally homogeneous regions method

In this work we solve the problem of automatic binary classification of subjects with a diagnosis of schizophrenia and control groups on a data set obtained on a Siemens 3T tomograph. The data set included 36 subjects undergoing treatment at Psychiatric Hospital no. 1 Named after N.A. Alexeev of the Department of Health of Moscow (GBUZ PKB No. 1 DZM) and 36 subjects from the control group. Machine learning methods were used to solve this problem. As a result, an accuracy of 76% was achieved, which corresponds to the results obtained in other scientific studies. The highest accuracy was obtained for the local homogeneity parameter (regional homogeneity – ReHo), already known in the literature. At the same time, the set of features developed by the authors based on the method for identifying functionally homogeneous regions (FHR) gave a classification accuracy of 74%. But at the same time, the set of FHR features provides higher classification accuracy when using a small number of brain regions. For example, already in 8 regions, the FHR set provided an almost maximum classification accuracy of 72.5% (versus 65% for the ReHo set), which suggests that it is the selected 8 regions that give the highest level of separation.

Limitations of Feature Attribution in Long Text Classification of Standards

Managing complex AI systems requires insight into a model's decision-making processes. Understanding how these systems arrive at their conclusions is essential for ensuring reliability. In the field of explainable natural language processing, many approaches have been developed and evaluated. However, experimental analysis of explainability for text classification has been largely constrained to short text and binary classification. In this applied work, we study explainability for a real-world task where the goal is to assess the technological suitability of standards. This prototypical use case is characterized by large documents, technical language, and a multi-label setting, making it a complex modeling challenge. We provide an analysis of approx. 1000 documents with human-annotated evidence. We then present experimental results with two explanation methods evaluating plausibility and runtime of explanations. We find that the average runtime for explanation generation is at least 5 minutes and that the model explanations do not overlap with the ground truth. These findings reveal limitations of current explanation methods. In a detailed discussion, we identify possible reasons and how to address them on three different dimensions: task, model and explanation method. We conclude with risks and recommendations for the use of feature attribution methods in similar settings.

Increase of Reliability of Anomalies Detection on Images at Formation of Their Feature Vectors in Wavelet Bases

The method of detection of life rafts and lifeboats in the water area of seas and oceans after shipwrecks based on the recognition of anomalies in the processed images, which increases the probability of recognition of monitoring objects, is proposed. The approach to solving such a problem is substantiated. The formulation of the problem of object recognition from the perspective of binary classification in the detection of anomalies is presented. The analytical expression for the decision-making algorithm is obtained. The possibility of formalization of image matrices in the form of histograms of color (brightness) intensity distributions is considered. The contrast of the feature space on their basis is estimated. It is suggested that the contrast of feature spaces be increased due to the secondary processing of histograms of distributions in the basis of multiple-scale wavelet decomposition. The possibility of realization of wavelet transformations on the basis of Haar functions and Gauss wavelets of the 1st and 2nd orders is considered. The mechanism of formation of secondary feature vectors from three-dimensional wavelet transforms by averaging their coefficients along the time shift axis is substantiated. It is shown that at the same dimensionality of histograms of brightness distribution with newly formed feature vectors, the latter provide higher contrast of feature spaces. It is recommended to use a Gaussian wavelet of the 2nd order for the formalization of images in jpeg format, which provides, other things being equal, a greater magnitude of differences for images containing anomalies. An approach to probabilistic evaluation of the algorithm for automatic image recognition is developed. The analytical expression is obtained and its constituent elements are justified. Graphical dependences of the probability of correct detection (recognition) of anomalies, depending on the size in relation to the total area of the frame and the dispersion of the underlying background are given. The results of the experiment on image recognition with a lifeboat in the ocean water area are presented. The directions of further research are defined.

Estimating the health of turbine engine based on the relationship between torque margin and density altitude

We present an anomaly detection method developed for the PHM North America 2024 Conference Data Challenge. This competition is aimed at estimating the health of helicopter turbine engines (PHM Society, 2024). The task includes the estimation of the torque margin (regression) and the health state (binary classification) of turbine engines. We developed an estimation model using a hybrid algorithm that combines data-based machine learning and domain knowledge-based processing. Our method achieved scores over 0.99 for both the testing and validation datasets. based on the calculation rules provided by PHM Society. These results were ranked first among all the participating teams.

Robust Health Condition Prediction of Helicopter Turboshaft Engines Using Ensemble Machine Learning Models

This paper presents a novel ensemble approach that combines multiple machine-learning algorithms to deliver robust predictions of helicopter turboshaft engine health status (nominal or faulty) using operational data. Engine health is evaluated through the torque margin, defined as the percentage difference between the measured and target torque values. A Gaussian process model is used to estimate the torque margin as a probability distribution function (PDF), and these predictions are incorporated as features into various machine-learning models. These models are then employed to perform binary classification, determining the engine's health state. To enhance performance, a reference set is defined for each unseen data point, allowing a comparison of the relative performances of the models, with the best performer selected for the final prediction. Our ensemble method achieves high accuracy in health classification while providing precise torque margin estimates. The results demonstrate that ensemble models offer superior generalization and reliability compared to individual machine-learning algorithms, especially when applied to complex, multivariate datasets like those from helicopter turboshaft engines.

Development of an automated artificial intelligence-based system for urogenital schistosomiasis diagnosis using digital image analysis techniques and a robotized microscope.

Urogenital schistosomiasis is considered a Neglected Tropical Disease (NTD) by the World Health Organization (WHO). It is estimated to affect 150 million people worldwide, with a high relevance in resource-poor settings of the African continent. The gold-standard diagnosis is still direct observation of Schistosoma haematobium eggs in urine samples by optical microscopy. Novel diagnostic techniques based on digital image analysis by Artificial Intelligence (AI) tools are a suitable alternative for schistosomiasis diagnosis. Digital images of 24 urine sediment samples were acquired in non-endemic settings. S. haematobium eggs were manually labeled in digital images by laboratory professionals and used for training YOLOv5 and YOLOv8 models, which would achieve automatic detection and localization of the eggs. Urine sediment images were also employed to perform binary classification of images to detect erythrocytes/leukocytes with the MobileNetv3Large, EfficientNetv2, and NasNetLarge models. A robotized microscope system was employed to automatically move the slide through the X-Y axis and to auto-focus the sample. A total number of 1189 labels were annotated in 1017 digital images from urine sediment samples. YOLOv5x training demonstrated a 99.3% precision, 99.4% recall, 99.3% F-score, and 99.4% mAP0.5 for S. haematobium detection. NasNetLarge has an 85.6% accuracy for erythrocyte/leukocyte detection with the test dataset. Convolutional neural network training and comparison demonstrated that YOLOv5x for the detection of eggs and NasNetLarge for the binary image classification to detect erythrocytes/leukocytes were the best options for our digital image database. The development of low-cost novel diagnostic techniques based on the detection and identification of S. haematobium eggs in urine by AI tools would be a suitable alternative to conventional microscopy in non-endemic settings. This technical proof-of-principle study allows laying the basis for improving the system, and optimizing its implementation in the laboratories.

Scalable deep learning artificial intelligence histopathology slide analysis and validation.

Deep learning involves an artificial intelligence (AI) approach and has been shown to provide superior performance for automating image recognition tasks, as well as exceeding human capabilities in both time and accuracy. Histopathology diagnostics is one of the more popular challenges at the intersection of artificial intelligence, computer vision, and medicine. Developing methods to automatically detect and identify pathologies in digitized histology slides imposes unique challenges due to the large size of these images and the complexity of the features present in biological tissue. Most methods that are capable of human-level recognition in histopathology are tuned to a specific problem since the computational complexity exceeds that of traditional image classification problems. In the current study, a deep learning approach is developed and presented that can be trained to locate and accurately classify different types of pathologies in gigapixel digitized histology slides along with completing the binary disease classification for the entire image. The approach uses a novel pyramid tiling approach to take advantage of spatial awareness around the area to be classified, while maintaining efficiency and scalability for gigapixel images. The approach is trained and validated on a wide variety of tissue types (i.e., testis, ovary, prostate, kidney) and pathologies taken from an epigenetically altered histology study at Washington State University. The newly developed procedure was optimized and validated along with comparison and validation on public histology datasets. The current developed procedure was found to be optimal and more reproducible when compared to manual procedures, and optimal to previous protocols that used fragmented tissue or slide analysis. Observations demonstrate that the deep learning histopathology analysis is significantly more efficient and accurate than standard manual histopathology analysis.

Positive Unlabeled Learning Selected Not At Random (PULSNAR): class proportion estimation without the selected completely at random assumption

Positive and unlabeled (PU) learning is a type of semi-supervised binary classification where the machine learning algorithm differentiates between a set of positive instances (labeled) and a set of both positive and negative instances (unlabeled). PU learning has broad applications in settings where confirmed negatives are unavailable or difficult to obtain, and there is value in discovering positives among the unlabeled (e.g., viable drugs among untested compounds). Most PU learning algorithms make the selected completely at random (SCAR) assumption, namely that positives are selected independently of their features. However, in many real-world applications, such as healthcare, positives are not SCAR (e.g., severe cases are more likely to be diagnosed), leading to a poor estimate of the proportion, α, of positives among unlabeled examples and poor model calibration, resulting in an uncertain decision threshold for selecting positives. PU learning algorithms vary; some estimate only the proportion, α, of positives in the unlabeled set, while others calculate the probability that each specific unlabeled instance is positive, and some can do both. We propose two PU learning algorithms to estimate α, calculate calibrated probabilities for PU instances, and improve classification metrics: i) PULSCAR (positive unlabeled learning selected completely at random), and ii) PULSNAR (positive unlabeled learning selected not at random). PULSNAR employs a divide-and-conquer approach to cluster SNAR positives into subtypes and estimates α for each subtype by applying PULSCAR to positives from each cluster and all unlabeled. In our experiments, PULSNAR outperformed state-of-the-art approaches on both synthetic and real-world benchmark datasets.

CSTAN: A Deepfake Detection Network with CST Attention for Superior Generalization

With the advancement of deepfake forgery technology, highly realistic fake faces have posed serious security risks to sensor-based facial recognition systems. Recent deepfake detection models mainly use binary classification models based on deep learning. Despite achieving high detection accuracy on intra-datasets, these models lack generalization ability when applied to cross-datasets. We propose a deepfake detection model named Channel-Spatial-Triplet Attention Network (CSTAN), which focuses on the difference between real and fake features, thereby enhancing the generality of the detection model. To enhance the feature-learning ability of the model for image forgery regions, we have designed the Channel-Spatial-Triplet (CST) attention mechanism, which extracts subtle local information by capturing feature channels and the spatial correlation of three different scales. Additionally, we propose a novel feature extraction method, OD-ResNet-34, by embedding ODConv into the feature extraction network to enhance its dynamic adaptability to data features. Trained on the FF++ dataset and tested on the Celeb-DF-v1 and Celeb-DF-v2 datasets, the experimental results show that our model has stronger generalization ability in cross-datasets than similar models.

A Field-Programmable Gate Array-Based Adaptive Sleep Posture Analysis Accelerator for Real-Time Monitoring

This research presents a sleep posture monitoring system designed to assist the elderly and patient attendees. Monitoring sleep posture in real time is challenging, and this approach introduces hardware-based edge computation methods. Initially, we detected the postures using minimally optimized sensing modules and fusion techniques. This was achieved based on subject (human) data at standard and adaptive levels using posture-learning processing elements (PEs). Intermittent posture evaluation was performed with respect to static and adaptive PEs. The final stage was accomplished using the learned subject posture data versus the real-time posture data using posture classification. An FPGA-based Hierarchical Binary Classifier (HBC) algorithm was developed to learn and evaluate sleep posture in real time. The IoT and display devices were used to communicate the monitored posture to attendant/support services. Posture learning and analysis were developed using customized, reconfigurable VLSI architectures for sensor fusion, control, and communication modules in static and adaptive scenarios. The proposed algorithms were coded in Verilog HDL, simulated, and synthesized using VIVADO 2017.3. A Zed Board-based field-programmable gate array (FPGA) Xilinx board was used for experimental validation.

Reliability Analysis of Complex Structures Under Multi-Failure Mode Utilizing an Adaptive AdaBoost Algorithm

A reliability analysis can become intricate when addressing issues related to nonlinear implicit models of complex structures. To improve the accuracy and efficiency of such reliability analyses, this paper presents a surrogate model based on an adaptive AdaBoost algorithm. This model employs an adaptive method to determine the optimal training sample set, ensuring it is as evenly distributed as possible on both sides of the failure curve and fully contains the information it represents. Subsequently, with the integration and iterative characteristics of the AdaBoost algorithm, a simple binary classifier is iteratively applied to build a high-precision alternative model for complex structural fault diagnosis to cope with multiple failure modes. Then, the Monte Carlo simulation technique is employed to meticulously assess the failure probability. The accuracy and stability of the proposed method’s iterative convergence process are validated through three numerical examples. The findings of the study illuminate that the proposed method is not only remarkably precise but also exceptionally efficient, capable of addressing the challenges related to the reliability evaluation of complex structures under multi-failure mode. The method proposed in this paper enhances the application of mechanical structures and facilitates the utilization of complex mechanical designs.

Distributed Estimation of Principal Support Vector Machines for Sufficient Dimension Reduction

The principal support vector machines method (Li et al., 2011) is a powerful tool for sufficient dimension reduction that replaces original predictors with their low-dimensional linear combinations while preserving the information for regression and classification. However, the computational burden of the principal support vector machines method constrains its use for massive data. To address this issue, we propose a naive and a refined distributed estimation algorithms for fast implementation when the sample size is large. Both distributed sufficient dimension reduction estimators exhibit the same statistical efficiency as when all the data is merged together, which provides rigorous statistical guarantees for their application to large-scale datasets, while the refined method requires smaller batch sample sizes and hence is more advantageous when memory limitations exist on distributed machines. The two distributed algorithms are further adapted to principal weighted support vector machines (Shin et al., 2017) for sufficient dimension reduction in binary classification. The statistical accuracy and computational complexity of our proposed methods are examined through comprehensive simulation studies and in a real data application with more than 600000 samples.

A Comparison of Uncertainty Estimation Methods for Building Footprint Change Detection from Sentinel-2 Imagery

Abstract. This manuscript investigates the effects of uncertainty methods applied to the problem of deep learning-based semantic segmentation of building footprints on moderate-resolution satellite imagery. While the recent efforts of big corporations to add information about building locations and sizes on a global or continental scale are generally valuable, still the overall challenge persists in identifying the spatial-temporal patterns of growing urbanization. In this work, we extend UNet-type architectures to perform binary building footprint classification based on Sentinel-2 imagery resulting in five different models. While previous studies focused on urban areas in the Western world we conduct all training and evaluation in India. All models are trained on Microsoft building footprint products while for evaluation purposes high-quality reference data is manually selected from regions with especially good open-street-map coverage. Quantitative and qualitative experiments are conducted where a significant performance gain is found for a model trained with a mixture of aleatoric and epistemic uncertainty measures. The performance gain is even more pronounced for subsequent quantitative multi-temporal change detection experiments.

Water status assessment in grapevines using plant electrophysiology

Traditional methods for assessing vine water status, such as the Scholander pressure chamber, are time-consuming, punctual and labour-intensive. The development of alternative methods which are accurate, reliable and can provide real-time information on vine water status is a necessity for farmers all over the world. This study proposes the use of plant electrophysiology as a novel approach for real-time water status assessment in grapevines. We conducted four climate chamber experiments with potted grapevines under different irrigation regimes. Various morphological and physiological assessments were performed in parallel with electrophysiological measurements to correlate classic water status assessment methods with plant electrophysiological signals. Two machine learning approaches based on classification and regression were employed to train the prediction models. Results obtained from both models indicate significant differences in irrigation status between well-watered and water-deficit plants, with the latter showing reduced growth and physiological activity, confirming the water stress status of the plant. While the binary classification model successfully differentiates between well-watered and water-deficit plants, its practical use is limited. Therefore, a regression model was developed to directly predict predawn leaf water potential. To the best of our knowledge, this is the first time that electrical signals are correlated with vine water potential measurements. The findings presented here thus provide a promising new tool for future real-time and remote monitoring of vine water status to manage irrigation and adapt agronomic strategies. Nevertheless, validation and optimisation of the models are still necessary, particularly under field conditions.

Use of ResNets for HLB Disease Detection on Orange Leaves Using Terrestrial Multispectral Images

Abstract. Huanglongbing (HLB) is a bacterial disease transmitted by different vectors of sap-sucking insects. It affects all crops of citrus trees, decreasing the values of those fruits in the market and eventually the decay of orchards. In Brazil, the world's leading orange producer, citriculture faces severe issues with HLB and substantial economic loss. Technical means of scanning the orchards with high-throughput becomes essential for the sustainability of this industry. In this study, we propose to investigate an operational strategy consisting of scanning large portions of foliage (the canopy of one tree or more) in which there can few early foliage symptoms. It is proposed to investigate deep learning tools to solve this complex binary classification problem. The study is based on a dataset comprising 1,297 terrestrial multispectral (14 channels) images captured at high spatial resolution in a commercial orange orchard in Brazil. It is proposed to adapt and retrain standard neural network architectures, namely ResNets18 and ResNets34, to process such images. Our analysis reveals promising results, with both models demonstrating convergence and achieving stable performance. Notably, ResNet18 outperformed ResNet34, achieving an accuracy of 76.45% compared to 66.79% from ResNet34. These findings suggest that deep neural network methods can effectively manage non-radiometrically calibrated data and accurately distinguish images with HLB symptoms from healthy plants . However, with reduced datasets and limited possibilities for transfer learning and fine-tuning, it seems that only reasonable sized networks can be trained. Thus, more advanced state-of-the-art tools of the are still challenging to deploy for agricultural multi- or hyperspectral data.

Salivary Antibody Responses to Potentially Waterborne and Environmentally Transmitted Infections Among Two Tribal Nations in the Southwest United States.

Tribal Nations disproportionately lack access to safe drinking water and can be adversely affected by other water quality and environmental concerns. Such conditions could lead to an increase in the transmission of waterborne, environmental and hygiene related infections. We collected saliva samples from attendees at two Tribal Nation annual festivals and tested them for salivary immunoglobulin G (IgG) responses to selected common infections using an in-house multiplex immunoassay. Antibody responses were compared to responses from a previously conducted study in the midwestern United States. We collected and tested 531 samples from Tribal Nation sites and used data on 453 previously analyzed samples from the Midwest site. Logistic and linear regression models were used to model a binary classification of seropositivity and the intensity of the antibody response, respectively. Seroprevalence of chronic infections (Helicobacter pylori and Toxoplasma gondii) were generally consistent with estimates from population-based studies. Compared to the Midwest site, one of the Tribal Nation sites had consistently higher median antibody responses to several noroviruses. The Tribal Nation sites had a lower seroprevalence of hepatitis E virus antibodies. At the Tribal Nation sites, farm residents had higher antibody responses to Cryptosporidium spp., bottled water consumption was associated with lower responses to Cryptosporidium spp., animal contact was associated with T. gondii seropositivity, and recent diarrhea was associated with higher norovirus antibody responses. Helicobacter pylori seropositivity was associated with reduced odds of reporting allergies. This study demonstrated the application of a multiplex salivary immunoassay in Tribal Nations to provide insights regarding selected common pathogens which are transmitted through different transmission pathways including person-to-person contacts, contaminated food, soil and drinking water.

Cumulative learning based segmentation aided cell mixtures classification in digital holographic microscopy

The tumor microenvironment consists of several components like tumor cells, normal healthy cells, extracellular matrix, secreted factors, and other non-cellular components. The interaction among these components is crucial for tumor progression and metastasis. Identifying the presence of tumor cells among the normal healthy cells during the initial stages of cancer will aid in its early detection and treatment. One of the main features that differentiates a cancer cell from a normal healthy cell distinctively is its morphology. In this study, the effectiveness of machine learning algorithms is evaluated in classifying cells co-cultured in a Petri dish. To achieve this, digital holographic microscopy is used to capture the quantitative phase images of human dermal fibroblast and A375 human melanoma cancer cells co-cultured at 1:1, 1:2, and 2:1 ratios. Cell segmentation was performed using a cumulative learning approach. Subsequently, morphological and phase-based features were extracted to train the machine learning algorithms for binary classification of cells. Phase-based features in addition to the morphological features were found to provide improved performance of the classifier. The classifier demonstrated an average F1-score of 0.9 during testing.

Detection of Diabetic Retinopathy Using VGG19 and ResNet 50 Models

Diabetic Retinopathy (DR) is an ocular condition that can manifest in individuals living with Diabetes Mellitus (DM). Retinal fundus examinations must be conducted on DM individuals as early identification and treatment of DR can reduce the risk of impaired vision or blindness. The manual diagnosis of DR conducted by eye-care professionals can be tedious and time-consuming, especially during mass screenings. Deep learning (DL) techniques are being used to provide automated diagnosis of DR. This study adopted two CNN (VGG 19 and ResNet50) models for the binary classification of DR (Non-referable DR and Referable DR). Both models were trained and validated with retinal fundus images from publicly available datasets. After training with Kaggle dataset, VGG 19 and ResNet50 models achieved accuracies of 94.3% and 96.9% respectively. For external validation, varying levels of accuracy, sensitivity and specificity were obtained for the two models on different datasets. The sensitivity of the VGG 19 model for the Messidor 2 dataset was 78.8% while the sensitivity of the ResNet50 model for the Indian Diabetic Retinopathy Image Dataset (IDRiD) was 85.7%. Findings in this study have shown that DR detection with deep learning techniques can serve as an assistive tool for eye-care professionals in the future.

Mitigating false negatives in imbalanced datasets: An ensemble approach

Imbalanced datasets present a challenge in machine learning, especially in binary classification scenarios where one class significantly outweighs the other. This imbalance often leads to models favoring the majority class, resulting in inadequate predictions for the minority class, specifically in false negatives. In response to this issue, this work introduces the MinFNR ensemble algorithm, designed to minimize False Negative Rates (FNR) in imbalanced datasets. The new approach strategically combines data-level, algorithmic-level, and hybrid-level approaches to enhance overall predictive capabilities while minimizing computational resources using the Set Covering Problem (SCP) formulation. Through a comprehensive evaluation of diverse datasets, MinFNR consistently outperforms individual algorithms, showing its potential for applications where the cost of false negatives is substantial, such as fraud detection and medical diagnosis. This work also contributes to ongoing efforts to improve the reliability and effectiveness of machine learning algorithms in real imbalanced scenarios.

HarDNet-based deep learning model for osteoporosis screening and bone mineral density inference from hand radiographs

PurposeOsteoporosis, affecting over 200 million individuals, often remains unrecognized and untreated, increasing the risk of fractures in older adults. Osteoporosis is typically diagnosed with bone mineral density (BMD) measured by dual-energy X-ray absorptiometry (DXA). This study aims to develop DeepDXA-Hand, a deep learning model using the efficient CNN-based deep learning architecture, for opportunistic osteoporosis screening from hand radiographs. MethodsDeepDXA-Hand utilizes a CNN-based, HarDNet, approach to predict BMD non-invasively. A total of 10,351 hand radiographs and DXA pairs were used for model training and validation. The model's interpretability was enhanced using GradCAM for hotspot analysis to determine the model's attention areas. ResultsThe predicted and ground truth BMD were significantly correlated with a correlation coefficient of 0.745. For binary classification of osteoporosis, DeepDXA-Hand demonstrated a sensitivity of 0.73, specificity of 0.83, and accuracy of 0.80, indicating its clinical potential. The model mainly focused on the carpal bones, such as the capitate, trapezoid, hamate, triquetrum, and the head of the second metacarpal bone, suggesting these areas provide radiological features for inferring BMD. ConclusionDeepDXA-Hand shows potential for the early detection of osteoporosis with high sensitivity and specificity. Further studies should explore its utility in predicting fracture risks. Mini abstractOsteoporosis affects millions and often goes undetected and untreated. DeepDXA-Hand, a HarDNet-based deep learning model, predicted bone mineral density with a correlation of 0.745 and classified osteoporosis with 0.80 accuracy. This model enhances early detection and has significant clinical potential as osteoporosis opportunistic screening tool.