Automatic Signal Classification Research Articles

CANova: A hybrid intrusion detection framework based on automatic signal classification for CAN

Over the years, vehicles have become increasingly complex and an attractive target for malicious adversaries. This raised the need for effective and efficient Intrusion Detection Systemss (IDSs) for onboard networks able to work with the stringent requirements and the heterogeneity of information transmitted on the Controller Area Network. While state-of-the-art solutions are effective in detecting specific types of anomalies and work on a subset of the CAN signals, no single method can perform better than the others on all types of attacks, particularly if they need to provide predictions to comply with the domain’s real-time constraints. In this paper, we present CANova, a modular framework that exploits the characteristics of the different Controller Area Network (CAN) packets to select the Intrusion Detection Systemss (IDSs) that better fits them. In particular, it uses flow- and payload-based IDSs to analyze the packets’ content and arrival time. We evaluate CANova by comparing its performance against state-of-the-art Intrusion Detection Systemss (IDSs) for in-vehicle network and a comprehensive set of synthetic and real attacks in real-world CAN datasets. We demonstrate that our approach can achieve good performances in terms of detection, false positive rates, and temporal performances.

Automatic defect depth estimation for ultrasonic testing in carbon fiber reinforced composites using deep learning

Ultrasonic testing (UT) is commonly used to inspect the geometric shape of internal damage in composite materials and the test results need to be interpreted by trained experts. In this work, an automatic signal classification method based on deep learning is proposed for depth estimation of the detects introduced by low-velocity impact (LVI) in carbon fiber reinforced plastics (CFRPs). Three kinds of neural networks, LSTM, CNN, and CNN-LSTM are used to analyze the attributes with different depths. Then, trained models are applied to identify the depth information of impact damage. The results show that the CNN-LSTM model is a more accurate in-depth classification for LVI defects in CFRP based on A-scan signals than the other two structures.

Automated Damage Detection Using Lamb Wave-Based Phase-Sensitive OTDR and Support Vector Machines.

In this paper, we propose and demonstrate a damage detection technique based on the automatic classification of the Lamb wave signals acquired on a metallic plate. In the reported experiments, Lamb waves are excited in an aluminum plate through a piezoelectric transducer glued onto the monitored structure. The response of the monitored structure is detected through a high-resolution phase-sensitive optical time-domain reflectometer (ϕ-OTDR). The presence and location of a small perturbation, induced by placing a lumped mass of 5 g on the plate, are determined by processing the optical fiber sensor data through support vector machine (SVM) classifiers trained with experimental data. The results show that the proposed method takes full advantage of the multipoint sensing nature of the ϕ-OTDR technology, resulting in accurate damage detection and localization.

A Study on Classification Method of Mine Vibration Based on Microseismic Monitoring Cloud Service Platform

With the gradual deepening of mining depth, sudden ground pressure disasters such as rock burst and collapse caused by deep high stress and high rock pressure are major hidden dangers affecting mine safety production. Microseismic monitoring technology has been widely applied in the field of mine ground pressure disaster warning. The existing microseismic monitoring system has some problems, such as inaccurate automatic recognition and classification of waveform signals, low quality and low efficiency of manual processing. In this paper, the automatic classification of vibration signals in the process of mining is studied, and the automatic classification results are uploaded to the cloud service platform in real time, which solves the technical bottleneck of the existing microseismic monitoring system. Meanwhile, the real-time monitoring of mine ground pressure safety is guaranteed based on the cloud service platform of microseismic monitoring.

ECG Signals Segmentation Using Deep Spatiotemporal Feature Fusion U-Net for QRS Complexes and R-Peak Detection

To detect and identify QRS complexes and R-peak is one of the crucial steps in the field of electrocardiogram (ECG) signals research, and their detection accuracy directly affects the performance of the subsequent ECG signal processing and analysis. However, the task involving the detection and identification of the ECG features becomes so complex for the traditional threshold-based detection methods. At present, researchers turn to deep learning approaches but little work has been done on the detection of spatiotemporal features of the QRS complexes. In this paper, for better QRS complexes and R-peak detection, we propose a novel method based on the improved U-net model, called ST-Res U-net. This uniquely designed component of the improved U-net model is made up of four levels of ST Block (the block extracting spatial-temporal features) and Res Path (the residual path). The entire framework contains three steps, data preprocessing dealing with denoising of the raw ECG signals, the key component ST-Res U-net dealing with the spatiotemporal feature extraction, and a threshold screening algorithm for locating the R-peaks. Our experiments are purposedly designed to test various combination structures of the ST blocks. The training and testing are from the MIT-BIH arrhythmia database (MITDB) and the CPSC2019 database. We adopt a commonly used set of evaluation criteria with the following experimental results: 99.76% and 90.01% for sensitivity, 99.87% and 93.5% for positive predictivity rate, 99.81% and 91.75% for F1 value, and 0.37% and 15.24% for detection error rate. The former numbers are for MITDB and the latter CPSC2019. Further, we test the proposed method on the PTB-XL database (without labeled R-peak positions) and the method can accurately locate QRS complexes and R-peaks The experimental results demonstrate that the proposed model and method is quite effective for the automatic classification and annotation of ECG signals and thus greatly improves the accuracy of diagnosis of arrhythmia diseases.

Siamese Convolutional Neural Network for Heartbeat Classification Using Limited 12-Lead ECG Datasets

The Electrocardiogram (ECG) is a low-cost exam commonly used to diagnose abnormalities in the cardiac cycle. Over the years, the scientific community has investigated the automatic classification of ECG signals driven by advanced Machine Learning (ML) techniques. Despite recent scientific advances, annotating large and diverse datasets to support the training of ML techniques is still very time-consuming and error-prone. Indeed, ML techniques whose training does not require extensive and well-annotated datasets are becoming even more prominent. Therefore, it is possible to correctly identify and classify abnormalities in the cardiac cycle (e.g., rare cardiologic disturbs) using limited data available in ECG datasets. However, the classification of heartbeats from digital tracings of ECG signals containing 12 leads from imbalanced datasets is challenging due to many existing heart diseases. This study investigates the few-shot learning paradigm based on Siamese Convolutional Neural Networks (SCNN), popular in imaging classification problems, to classify 12-Lead ECG heartbeats using a few training samples with supervised information. The proposed SCNN model presented an accuracy of up to 95% in a public dataset based on the hold-out validation method, implemented for different combinations of similarity and loss functions. Besides, using the 7-fold cross-validation method, the model presented a mean area under the curve of 89%. We also compared the class-by-class classification results with those of similar methods available in the literature, obtaining the same or better results based on performance metrics such as accuracy, precision, recall, and specificity.

G2-ResNeXt: A Novel Model for ECG Signal Classification

Electrocardiograms (ECG) are the primary basis for the diagnosis of cardiovascular diseases. However, due to the large volume of patients’ ECG data, manual diagnosis is time-consuming and laborious. Therefore, intelligent automatic ECG signal classification is an important technique for overcoming the shortage of medical resources. This paper proposes a novel model for inter-patient heartbeat classification, named G2-ResNeXt, which adds a two-fold grouping convolution (G2) to the original ResNeXt structure, as to achieve better automatic feature extraction and classification of ECG signals. Experiments, conducted on the MIT-BIH arrhythmia database, confirm that the proposed model outperforms all state-of-the-art models considered (except the GRNN model for one of the heartbeat classes), by achieving <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">overall accuracy</i> of 96.16%, and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sensitivity</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">precision</i> of 97.09% and 95.90%, respectively, for the ventricular ectopic heartbeats (VEB), and of 80.59% and 82.26%, respectively, for the supraventricular ectopic heartbeats (SVEB).

Domain Knowledge Informed Multitask Learning for Landslide-Induced Seismic Classification

Automatic seismic signal classification methods are extensively investigated to reduce or replace manual interpretation, with great potential in previous research. Discriminative seismic wave propagation physical characteristics, such as velocities and accelerations, are rarely considered for classification. A multitask learning scheme is proposed that utilises the seismic wave equation and three-dimensional (3D) P-wave velocity <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Vp</i> model for signal representation learning. The classifier uses the obtained latent feature maps on a convolutional neural network architecture for classification of rockfall, slide quake, earthquake, and natural/anthropogenic noise events, recorded at an ongoing landslide. Our experimental results show that our approach outperforms state-of-the-art methods.

Первый сравнительный анализ метеорного эхо и спорадического рассеяния, идентифицированных самообучившейся нейронной сетью по данным радаров EKB и MAGW ИСЗФ СО РАН

The paper describes the current version (v.1.1) of the algorithm for automatic classification of signals received by ISTP SB RAS decameter coherent scatter radars. The algorithm is a self-learning neural network that determines the type of scattered signals from the results of physical modeling of radio wave propagation, using radar data and international reference models of the ionosphere and geomagnetic field. According to MAGW and EKB ISTP SB RAS radar data for 2021, the algorithm self-learns to classify scattered signals into initially unknown classes based on physically interpreted parameters of radio wave propagation and data measured by the radar, with 15 frequently observed out of 20 possible hidden classes identified, 14 of which can be interpreted from a physical point of view. To demonstrate the operation of the algorithm, we present the first statistical analysis of observations of signals assigned by the algorithm to classes which we interpret as scattering by meteor trails and scattering with the sporadic E layer respectively. Through a statistical analysis of EKB and MAGW radar data during 2021–2022, we demonstrate the range-altitude characteristics of signals of these types. A correlation is shown between the hourly average numbers of observations of both classes, as well as between the hourly average line-of-sight velocities obtained for both classes. The results obtained make it possible to interpret these classes as a meteor echo and sporadic scattering respectively, and to use radar data to study the interaction between the neutral atmosphere (studied from meteor scattering data) and the lower ionosphere (studied from observations of sporadic scattering). Currently, this classification algorithm works in ISTP SB RAS radars in automatic mode.

The first comparative analysis of meteor echo and sporadic scattering identified by a self-learned neural network in EKB and MAGW ISTP SB RAS radar data

The paper describes the current version (v.1.1) of the algorithm for automatic classification of signals received by ISTP SB RAS decameter coherent scatter radars. The algorithm is a self-learning neural network that determines the type of scattered signals from the results of physical modeling of radio wave propagation, using radar data and international reference models of the ionosphere and geomagnetic field. According to MAGW and EKB ISTP SB RAS radar data for 2021, the algorithm self-learns to classify scattered signals into initially unknown classes based on physically interpreted parameters of radio wave propagation and data measured by the radar, with 15 frequently observed out of 20 possible hidden classes identified, 14 of which can be interpreted from a physical point of view. To demonstrate the operation of the algorithm, we present the first statistical analysis of observations of signals assigned by the algorithm to classes which we interpret as scattering by meteor trails and scattering with the sporadic E layer respectively. Through a statistical analysis of EKB and MAGW radar data during 2021–2022, we demonstrate the range-altitude characteristics of signals of these types. A correlation is shown between the hourly average numbers of observations of both classes, as well as between the hourly average line-of-sight velocities obtained for both classes. The results obtained make it possible to interpret these classes as a meteor echo and sporadic scattering respectively, and to use radar data to study the interaction between the neutral atmosphere (studied from meteor scattering data) and the lower ionosphere (studied from observations of sporadic scattering). Currently, this classification algorithm works in ISTP SB RAS radars in automatic mode.

The use of weighted self-organizing maps to interrogate large seismic data sets

SUMMARY Modern microseismic monitoring systems can generate extremely large data sets with signals originating from a variety of natural and anthropogenic sources. These data sets may contain multiple signal types that require classification, analysis and interpretation: a considerable task if done manually. Machine learning techniques may be applied to these data sets to expedite and improve such analysis. In this study, we apply an unsupervised technique, the Self-Organizing Map (SOM), to high-volume data recorded by an in-mine microseismic network. This represents a good example of a large seismic data set that contains a wide range of signals, owing to the diversity of source processes occurring within the mine. The signals are quantified by extracting a number of features (temporal and spectral) from the waveforms which are provided as input data for the SOM. We develop and implement a weighted variant of the SOM in which the contributions of various different features to the training of the map are allowed to evolve. The standard and weighted SOMs are applied to the data, and the output maps compared. Both variants are able to separate source types based on the waveform characteristics, allowing for rapid, automatic classification of signals and the ability to find sources with similar waveforms. Fast classification of such signals provides practical benefit by automatically discarding waveforms associated with anthropogenic sources within the mine while seismic signals originating from genuine microseismic events, which constitute a small fraction of all signals, can be prioritized for subsequent processing and analysis. The weighted variant provides an exploratory tool through quantification of the contribution of different features to the clustering process. This helps to optimize the performance of the SOM through the identification of redundant features. Furthermore, those features that are assigned large weights are considered to be more representative of the source generation processes as they contribute more to the cluster separation process. We apply weighted SOMs to data from a mine recorded during two different time periods, corresponding to different stages of the mine development. Changes in feature importance and in the observed distribution of feature values indicate evolving source generation processes and may be used to support investigatory analysis. The weighted SOM therefore represents an effective tool to help manage and investigate large seismic data sets, providing both practical benefit and insight into underlying event mechanisms.

Self-trained artificial neural network for physical classification of ionospheric radar data

In the paper an algorithm for automatic classification of signals received by decameter coherent scatter radars is described. The algorithm is self-trained on unlabeled dataset and is based on training optimal physics-based classification using clusterization results. The approach allows to find optimal (hidden) classification, based on physically interpretable parameters of the received data, both obtained during physics-based numerical simulation and measured experimentally. As a result we trained the classification artificial neural network and found 11 well-interpretable classes from physical point of view in the available data. Other found classes are not interpretable from physical point of view, demonstrating the importance of taking into account radiowave propagation for correct classification. The advantages of the method include the possibility of its automatic training: it does not require preliminary data analysis by an expert, and it is self-trained to separate available data into the classes, that can be interpreted from a physical point of view.

Fully Automatic Classification of Cardiotocographic Signals with 1D-CNN and Bi-directional GRU.

Prenatal fetal monitoring, which can monitor the growth and health of the fetus, is vital for pregnant women before delivery. During pregnancy, it is essential to classify whether the fetus is abnormal, which helps physicians carry out early intervention to avoid fetal heart hypoxia and even death. Fetal heart rate and uterine contraction signals obtained by fetal heart monitoring equipment are essential to estimate fetal health status. In this paper, we pre-process the obtained data set and enhance them using Hermite interpolation on the abnormal classification in the samples. We use the 1D-CNN and GRU hybrid models to extract the abstract features of fetal heart rate and uterine contraction signals. Several evaluation metrics are used for evaluation, and the accuracy is 96 %, while the sensitivity is 95 %, and the specificity is 96 %. The experiments show the effectiveness of the proposed method, which can provide physicians and users with more stable, efficient, and convenient diagnosis and decision support.

Construction of Vocal Timbre Evaluation System Based on Classification Algorithm

With the continuous development of communication technology, computer technology, and network technology, a large amount of information such as images, videos, and audios has grown exponentially, and people have started to be exposed to massive multimedia contents, which can easily and quickly access the increasingly rich music resources, so new technologies are urgently needed for their effective management, and automatic classification of audio signals has become the focus of engineering and academic attention. Currently, music retrieval can be achieved by selecting song titles and singer names, but as people’s living standards continue to improve, the spiritual realm is also enriched. People want to be able to select music with different types of emotional expressions with their emotions. It mainly includes the basic principles of audio classification, the analysis and extraction of music emotion features, and the selection of the best classifier. Two classification algorithms, hybrid Gaussian model and AdaBoost, are used to classify music emotions, and the two classifiers are combined. In this paper, we propose the Discrete Harmonic Transform (DHT), a sparse transform based on harmonic frequencies. This paper derives and proves the formula of Discrete Harmonic Transform and further analyzes the harmonic structure of musical tone signal and the accuracy of harmonic structure. Since the timbre of musical instruments depends on the harmonic structure, and similar instruments have similar harmonic structures, the discrete harmonic transform coefficients can be defined as objective indicators corresponding to the timbre of musical instruments, and thus the concept of timbre expression spectrum is proposed, and a specific construction algorithm is given in this paper. In the application of musical instrument recognition, the 53-dimensional combined features of LPCC, MFCC, and timbre expression spectrum are selected, and a nonlinear support vector machine is used as the classifier. The classification recognition rate is improved by reducing the number of feature dimensions.

An automatic single beat algorithm to discriminate farfield from nearfield bipolar voltage electrograms from the pulmonary veins

Abstract Funding Acknowledgements Type of funding sources: Public hospital(s). Main funding source(s): “Stiftung für Herzschrittmacher und Elektrophysiologie” Basel, Switzerland Background Confirmation of pulmonary vein (PV) isolation (PVI) during ablation of atrial fibrillation can be challenging due to superimposition of nearfield (NF) PV and farfield (FF) atrial bipolar voltage electrograms (BVE). Purpose To develop an automatic algorithm allowing to discriminate PV nearfield (PV-NF) from atrial farfield (atrial-FF) BVE from a circular mapping catheter during cryoballoon (CB) PVI based on a single-heartbeat analysis. Methods BVEs from a decapolar inner-lumen diagnostic catheter (Achieve, Medtronic) during CB PVI were manually classified as PV-NF, atrial-FF and combined FF-NF signal based on the characteristics and disappearance of the PV signal during isolation (Figure, upper row). BVE power spectra were computed using the fast Fourier transform (FFT) and the automatic classification of PV-NF, atrial-FF and combined FF-NF signals was performed using the power in different frequency bands (Figure, lower row). Support vector machine classifier was used to identify PV-NF BVE due to its highest predictive accuracy for the two classes PV-NF+ (PV-NF only and combined FF-NF) and PV-NF- (atrial-FF only). Validation of the approach was performed by comparison of a subset of 80 random samples, which were classified in addition by five experienced electrophysiologists. Results We analysed a dataset of 355 BVEs from 57 patients. The examples were balanced between the two classes PV-NF+ and PV-NF-. The mean duration (95% CI) of the BVE was 58 ms (26 to 86), 70 ms (50 to 100) and 94 ms (71 to 139) for PV NF, atrial-FF and combined FF-NF, respectively. The overall balanced accuracy including BVE from all PVs was 82.7% (95% CI: 80.3% to 85.1%). The analysis on individual PVs showed an accuracy of 96.6%, 85.2%, 80.8%, and 76.9% for the right inferior, right superior, left inferior and left superior PV, respectively. Validation of the algorithm in the subset of 80 patients showed a comparable accuracy, sensitivity and specificity in PV-NF detection between the automatic algorithm and the experienced electrophysiologists (82.8%, 89.2%, and 76.3%, compared to 85.2%, 91.9%, and 78.5%, respectively). Conclusion A reliable automatic based classification algorithm to identify PV-NF BVE could be developed based on a single-beat analysis. Real-time applications as well as using other electrode configurations may improve local signal interpretation.

Classification of Epileptic Brain Signals of College Students Based on Deep Learning

Epilepsy, as a common nervous system disease in the world, urgently needs effective methods for diagnosis and treatment. Scalp electroencephalogram (EEG) contains a lot of physiological and pathological information, which plays a very important role in the diagnosis of brain diseases such as epilepsy. At present, the clinical analysis of EEG signal is mainly based on the clinician’s visual analysis, which makes the clinician’s task heavy, and there is no quantitative standard for the analysis results. Therefore, the automatic classification of epileptic EEG signals has great potential in current clinical application. Based on the deep learning VGG16 convolutional neural network model, a method for automatic detection of seizure signals is proposed in this paper. Firstly, the samples are preprocessed, then enter the VGG16 convolutional neural network model for training, and finally input the validation set for validation. The experimental results indicate that this model can distinguish the pre-ictal EEG, interictal EEG and ictal EEG.

Classification of EEG Signal-Based Encephalon Magnetic Signs for Identification of Epilepsy-Based Neurological Disorder.

Magnetoencephalography (MEG) is now widely used in clinical examinations and medical research in many fields. Resting-state magnetoencephalography-based brain network analysis can be used to study the physiological or pathological mechanisms of the brain. Furthermore, magnetoencephalography analysis has a significant reference value for the diagnosis of epilepsy. The scope of the proposed research is that this research demonstrates how to locate spikes in the phase locking functional brain connectivity network of the Desikan-Killiany brain region division using a neural network approach. It also improves detection accuracy and reduces missed and false detection rates. The automatic classification of epilepsy encephalomagnetic signals can make timely judgments on the patient's condition, which is of tremendous clinical significance. The existing literature's research on the automatic type of epilepsy EEG signals is relatively sufficient, but the research on epilepsy EEG signals is relatively weak. A full-band machine learning automatic discrimination method of epilepsy brain magnetic spikes based on the brain functional connection network is proposed. The four classifiers are comprehensively compared. The classifier with the best effect is selected, and the discrimination accuracy can reach 93.8%. Therefore, this method has a good application prospect in automatically identifying and labeling epileptic spikes in magnetoencephalography.

Automatic Multichannel Electrocardiogram Record Classification Using XGBoost Fusion Model.

There is an increasing demand for automatic classification of standard 12-lead electrocardiogram signals in the medical field. Considering that different channels and temporal segments of a feature map extracted from the 12-lead electrocardiogram record contribute differently to cardiac arrhythmia detection, and to the classification performance, we propose a 12-lead electrocardiogram signal automatic classification model based on model fusion (CBi-DF-XGBoost) to focus on representative features along both the spatial and temporal axes. The algorithm extracts local features through a convolutional neural network and then extracts temporal features through bi-directional long short-term memory. Finally, eXtreme Gradient Boosting (XGBoost) is used to fuse the 12-lead models and domain-specific features to obtain the classification results. The 5-fold cross-validation results show that in classifying nine categories of electrocardiogram signals, the macro-average accuracy of the fusion model is 0.968, the macro-average recall rate is 0.814, the macro-average precision is 0.857, the macro-average F1 score is 0.825, and the micro-average area under the curve is 0.919. Similar experiments with some common network structures and other advanced electrocardiogram classification algorithms show that the proposed model performs favourably against other counterparts in F1 score. We also conducted ablation studies to verify the effect of the complementary information from the 12 leads and the auxiliary information of domain-specific features on the classification performance of the model. We demonstrated the feasibility and effectiveness of the XGBoost-based fusion model to classify 12-lead electrocardiogram records into nine common heart rhythms. These findings may have clinical importance for the early diagnosis of arrhythmia and incite further research. In addition, the proposed multichannel feature fusion algorithm can be applied to other similar physiological signal analyses and processing.

Special issue editorial on emerging trends in internet of things for e‐health and medical supply chain systems

Special issue editorial on emerging trends in internet of things for e‐health and medical supply chain systems

Graphical representation learning-based approach for automatic classification of electroencephalogram signals in depression

Depression is a major depressive disorder characterized by persistent sadness and a sense of worthlessness, as well as a loss of interest in pleasurable activities, which leads to a variety of physical and emotional problems. It is a worldwide illness that affects millions of people and should be detected at an early stage to prevent negative effects on an individual's life. Electroencephalogram (EEG) is a non-invasive technique for detecting depression that analyses brain signals to determine the current mental state of depressed subjects. In this study, we propose a method for automatic feature extraction to detect depression by first constructing a graph from the dataset where the nodes represent the subjects in the dataset and where the edge weights obtained using the Euclidean distance reflect the relationship between them. The Node2vec algorithmic framework is then used to compute feature representations for nodes in a graph in the form of node embeddings ensuring that similar nodes in the graph remain near in the embedding. These node embeddings act as useful features which can be directly used by classification algorithms to determine whether a subject is depressed thus reducing the effort required for manual handcrafted feature extraction. To combine the features collected from the multiple channels of the EEG data, the method proposes three types of fusion methods: graph-level fusion, feature-level fusion, and decision-level fusion. The proposed method is tested on three publicly available datasets with 3, 20, and 128 channels, respectively, and compared to five state-of-the-art methods. The results show that the proposed method detects depression effectively with a peak accuracy of 0.933 in decision-level fusion, which is the highest among the state-of-the-art methods.