Murmur identification and outcome prediction in phonocardiograms using deep features based on Stockwell transform

Traditionally, heart murmurs are diagnosed through cardiac auscultation, which requires specialized training and experience. The purpose of this study is to predict patients' clinical outcomes (normal or abnormal) and identify the presence or absence of heart murmurs using phonocardiograms (PCGs) obtained at different auscultation points. A semi-supervised model tailored to PCG classification is introduced in this study, with the goal of improving performance using time-frequency deep features. The study begins by investigating the behavior of PCGs in the time-frequency domain, utilizing the Stockwell transform to convert the PCG signal into two-dimensional time-frequency maps (TFMs). A deep network named AlexNet is then used to derive deep feature sets from these TFMs. In feature reduction, redundancy is eliminated and the number of deep features is reduced to streamline the feature set. The effectiveness of the extracted features is evaluated using three different classifiers using the CinC/Physionet challenge 2022 dataset. For Task I, which focuses on heart murmur detection, the proposed approach achieved an average accuracy of 93%, sensitivity of 91%, and F1-score of 91%. According to Task II of the CinC/Physionet challenge 2022, the approach showed a clinical outcome cost of 5290, exceeding the benchmark set by leading methods in the challenge.

STNet: A Time-Frequency Analysis-Based Intrusion Detection Network for Distributed Optical Fiber Acoustic Sensing Systems

Distributed optical fiber acoustic sensing (DAS) is promising for long-distance intrusion-anomaly detection tasks. However, realistic settings suffer from high-intensity interference noise, compromising the detection performance of DAS systems. To address this issue, we propose STNet, an intrusion detection network based on the Stockwell transform (S-transform), for DAS systems, considering the advantages of the S-transform in terms of noise resistance and ability to detect disturbances. Specifically, the signal detected by a DAS system is divided into space-time data matrices using a sliding window. Subsequently, the S-transform extracts the time-frequency features channel by channel. The extracted features are combined into a multi-channel time-frequency feature matrix and presented to STNet. Finally, a non-maximum suppression algorithm (NMS), suitable for locating intrusions, is used for the post-processing of the detection results. To evaluate the effectiveness of the proposed method, experiments were conducted using a realistic high-speed railway environment with high-intensity noise. The experimental results validated the satisfactory performance of the proposed method. Thus, the proposed method offers an effective solution for achieving high intrusion detection rates and low false alarm rates in complex environments.

Fault Diagnosis in Centrifugal Pumps: A Dual-Scalogram Approach with Convolution Autoencoder and Artificial Neural Network

This paper proposes a new fault diagnosis method for centrifugal pumps by combining signal processing with deep learning techniques. Centrifugal pumps facilitate fluid transport through the energy generated by the impeller. Throughout the operation, variations in the fluid pressure at the pump's inlet may impact the generalization of traditional machine learning models trained on raw statistical features. To address this concern, first, vibration signals are collected from centrifugal pumps, followed by the application of a lowpass filter to isolate frequencies indicative of faults. These signals are then subjected to a continuous wavelet transform and Stockwell transform, generating two distinct time-frequency scalograms. The Sobel filter is employed to further highlight essential features within these scalograms. For feature extraction, this approach employs two parallel convolutional autoencoders, each tailored for a specific scalogram type. Subsequently, extracted features are merged into a unified feature pool, which forms the basis for training a two-layer artificial neural network, with the aim of achieving accurate fault classification. The proposed method is validated using three distinct datasets obtained from the centrifugal pump under varying inlet fluid pressures. The results demonstrate classification accuracies of 100%, 99.2%, and 98.8% for each dataset, surpassing the accuracies achieved by the reference comparison methods.

Quantized Information in Spectral Cyberspace

The constant-Q Gabor atom is developed for spectral power, information, and uncertainty quantification from time-frequency representations. Stable multiresolution spectral entropy algorithms are constructed with continuous wavelet and Stockwell transforms. The recommended processing and scaling method will depend on the signature of interest, the desired information, and the acceptable levels of uncertainty of signal and noise features. Selected Lamb wave signatures and information spectra from the 2022 Tonga eruption are presented as representative case studies. Resilient transformations from physical to information metrics are provided for sensor-agnostic signal processing, pattern recognition, and machine learning applications.

[Construction of an epileptic seizure prediction model using a semi-supervised method of generative adversarial and long short term memory network combined with Stockwell transform]

OBJECTIVE

To propose a semi-supervised epileptic seizure prediction model (ST-WGAN-GP-Bi-LSTM) to enhance the prediction performance by improving time-frequency analysis of electroencephalogram (EEG) signals, enhancing the stability of the unsupervised feature learning model and improving the design of back-end classifier.

METHODS

Stockwell transform (ST) of the epileptic EEG signals was performed to locate the time-frequency information by adaptive adjustment of the resolution and retaining the absolute phase to obtain the time-frequency inputs. When there was no overlap between the generated data distribution and the real EEG data distribution, to avoid failure of feature learning due to a constant JS divergence, Wasserstein GAN was used as a feature learning model, and the cost function based on EM distance and gradient penalty strategy was adopted to constrain the unsupervised training process to allow the generation of a high-order feature extractor. A temporal prediction model was finally constructed based on a bi-directional long short term memory network (Bi-LSTM), and the classification performance was improved by obtaining the temporal correlation between high-order time-frequency features. The CHB-MIT scalp EEG dataset was used to validate the proposed patient-specific seizure prediction method.

RESULTS

The AUC, sensitivity, and specificity of the proposed method reached 90.40%, 83.62%, and 86.69%, respectively. Compared with the existing semi-supervised methods, the propose method improved the original performance by 17.77%, 15.41%, and 53.66%. The performance of this method was comparable to that of a supervised prediction model based on CNN.

CONCLUSION

The utilization of ST, WGAN-GP, and Bi-LSTM effectively improves the prediction performance of the semi-supervised deep learning model, which can be used for optimization of unsupervised feature extraction in epileptic seizure prediction.

ECG quality assessment based on hand-crafted statistics and deep-learned S-transform spectrogram features

Background and Objective Electrocardiogram (ECG) quality assessment is significant for automatic diagnosis of cardiovascular disease and reducing the massive workload of reviewing continuous ECGs. Hence, how to design an appropriate algorithm for objectively evaluating the multi-lead ECG recordings is particularly important. Despite the deep learning methods performing well in many fields, as a data-driven method, it may not be entirely suitable for ECG analysis due to the difficulty in obtaining sufficient data and the low signal-to-noise ratio of ECG recordings. In this study, with the aim of providing an accurate and automatic ECG quality assessment scheme, we propose an innovative ECG quality assessment algorithm based on hand-crafted statistical features and deep-learned spectral features. Methods In this paper, a novel approach, combining the deep-learned Stockwell transform (S-Transform) spectrogram features and hand-crafted statistical features, is proposed for ECG quality assessment. Firstly, a double-input convolutional neural network (CNN) is established. Then, the S-Transform with a novel online augmentation scheme is performed on the multi-lead raw ECG signal received from one input layer to obtain proper time-frequency representation. After that, the CNN with three convolutional layers is employed to extract robust deep-learned features automatically. Simultaneously, the hand-crafted statistical features, including lead-fall, baseline drift, and R peak features, are calculated and fed into another input layer for feature fusion training. Finally, the deep-learned and hand-crafted features are concatenated and further fused by a fully connected layer for quality classification. Furthermore, a log-odds analysis scheme combining with a gradient-based method can localize the abnormal zone in time, frequency, and spatial domains. Results and Conclusion Our proposed method is evaluated on a publicly available database with 10-fold cross-validation. The experimental results demonstrate that the proposed assessment algorithm reached a mean accuracy of 93.09%, a mean F1-score of 0.8472, and a sensitivity of 0.9767. Moreover, comprehensive experiments indicate that the fusion of CNN features and statistical features has complementary advantages and ideal interpretability, achieving end-to-end multi-lead ECG assessment with satisfying performance.

An Explainable AI-Based Fault Diagnosis Model for Bearings

In this paper, an explainable AI-based fault diagnosis model for bearings is proposed with five stages, i.e., (1) a data preprocessing method based on the Stockwell Transformation Coefficient (STC) is proposed to analyze the vibration signals for variable speed and load conditions, (2) a statistical feature extraction method is introduced to capture the significance from the invariant pattern of the analyzed data by STC, (3) an explainable feature selection process is proposed by introducing a wrapper-based feature selector-Boruta, (4) a feature filtration method is considered on the top of the feature selector to avoid the multicollinearity problem, and finally, (5) an additive Shapley explanation followed by k-NN is proposed to diagnose and to explain the individual decision of the k-NN classifier for debugging the performance of the diagnosis model. Thus, the idea of explainability is introduced for the first time in the field of bearing fault diagnosis in two steps: (a) incorporating explainability to the feature selection process, and (b) interpretation of the classifier performance with respect to the selected features. The effectiveness of the proposed model is demonstrated on two different datasets obtained from separate bearing testbeds. Lastly, an assessment of several state-of-the-art fault diagnosis algorithms in rotating machinery is included.

A skew-based method for identifying intracranial EEG channels with epileptic activity without detecting spikes, ripples, or fast ripples

OBJECTIVE

To develop a method for identifying intracranial EEG (iEEG) channels with epileptic activity without the need to detect spikes, ripples, or fast ripples.

METHODS

We compared the skew of the distribution of power values from five minutes non-rapid eye movement stage N3 sleep for the 5-80 Hz, 80-250 Hz (ripple), and 250-500 Hz (fast ripple) bands of epileptic (located in seizure-onset or irritative zone) and non-epileptic iEEG channels recorded in patients with drug-resistant focal epilepsy. We optimized settings in 120 bipolar channels from 10 patients, compared the results to 120 channels from another 10 patients, and applied the method to channels of 12 individual patients.

RESULTS

The distribution of power values was more skewed in epileptic than in non-epileptic channels in all three frequency bands. The differences in skew were correlated with the presence of spikes, ripples, and fast ripples. When classifying epileptic and non-epileptic channels, the mean accuracy over 12 patients was 0.82 (sensitivity: 0.76, specificity: 0.91).

CONCLUSIONS

The 'skew method' can distinguish epileptic from non-epileptic channels with good accuracy and, in particular, high specificity.

SIGNIFICANCE

This is an easy-to-apply method that circumvents the need to visually mark or automatically detect interictal epileptic events.

SEMG Feature Extraction Based on StockwellTransform Improves Hand MovementRecognition Accuracy

Feature extraction, as an important method for extracting useful information from surface electromyography (SEMG), can significantly improve pattern recognition accuracy. Time and frequency analysis methods have been widely used for feature extraction, but these methods analyze SEMG signals only from the time or frequency domain. Recent studies have shown that feature extraction based on time-frequency analysis methods can extract more useful information from SEMG signals. This paper proposes a novel time-frequency analysis method based on the Stockwell transform (S-transform) to improve hand movement recognition accuracy from forearm SEMG signals. First, the time-frequency analysis method, S-transform, is used for extracting a feature vector from forearm SEMG signals. Second, to reduce the amount of calculations and improve the running speed of the classifier, principal component analysis (PCA) is used for dimensionality reduction of the feature vector. Finally, an artificial neural network (ANN)-based multilayer perceptron (MLP) is used for recognizing hand movements. Experimental results show that the proposed feature extraction based on the S-transform analysis method can improve the class separability and hand movement recognition accuracy compared with wavelet transform and power spectral density methods.

Automated detection of congestive heart failure from electrocardiogram signal using Stockwell transform and hybrid classification scheme

BACKGROUND AND OBJECTIVE

The congestive heart failure (CHF) is a life-threatening cardiac disease which arises when the pumping action of the heart is less than that of the normal case. This paper proposes a novel approach to design a classifier-based system for the automated detection of CHF.

METHODS

The approach is founded on the use of the Stockwell (S)-transform and frequency division to analyze the time-frequency sub-band matrices stemming from electrocardiogram (ECG) signals. Then, the entropy features are evaluated from the sub-band matrices of ECG. A hybrid classification scheme is adopted taking the sparse representation classifier and the average of the distances from the nearest neighbors into account for the detection of CHF. The proposition is validated using ECG signals from CHF subjects and normal sinus rhythm from public databases.

RESULTS

The results reveal that the proposed system is successful for the detection of CHF with an accuracy, a sensitivity and a specificity values of 98.78%, 98.48%, and 99.09%, respectively. A comparison with the existing approaches for the detection of CHF is accomplished.

CONCLUSIONS

The time-frequency entropy features of the ECG signal in the frequency range from 11 Hz to 30 Hz have higher performance for the detection of CHF using a hybrid classifier. The approach can be used for the automated detection of CHF in tele-healthcare monitoring systems.

Use of the Stockwell Transform in the Detection of P300 Evoked Potentials with Low-Cost Brain Sensors

The evoked potential is a neuronal activity that originates when a stimulus is presented. To achieve its detection, various techniques of brain signal processing can be used. One of the most studied evoked potentials is the P300 brain wave, which usually appears between 300 and 500 ms after the stimulus. Currently, the detection of P300 evoked potentials is of great importance due to its unique properties that allow the development of applications such as spellers, lie detectors, and diagnosis of psychiatric disorders. The present study was developed to demonstrate the usefulness of the Stockwell transform in the process of identifying P300 evoked potentials using a low-cost electroencephalography (EEG) device with only two brain sensors. The acquisition of signals was carried out using the Emotiv EPOC^® device—a wireless EEG headset. In the feature extraction, the Stockwell transform was used to obtain time-frequency information. The algorithms of linear discriminant analysis and a support vector machine were used in the classification process. The experiments were carried out with 10 participants; men with an average age of 25.3 years in good health. In general, a good performance (75⁻92%) was obtained in identifying P300 evoked potentials.

Application of the Stockwell Transform to Electroencephalographic Signal Analysis during Gait Cycle

The analysis of electroencephalographic signals in frequency is usually not performed by transforms that can extract the instantaneous characteristics of the signal. However, the non-steady state nature of these low voltage electrical signals makes them suitable for this kind of analysis. In this paper a novel tool based on Stockwell transform is tested, and compared with techniques such as Hilbert-Huang transform and Fast Fourier Transform, for several healthy individuals and patients that suffer from lower limb disability. Methods are compared with the Weighted Discriminator, a recently developed comparison index. The tool developed can improve the rehabilitation process associated with lower limb exoskeletons with the help of a Brain-Machine Interface.

Automatic seizure detection using Stockwell transform and boosting algorithm for long-term EEG

Automatic detection of seizures has vital significance for epileptic diagnosis and can efficiently reduce the workload of the medical staff. In this study, a novel seizure detection method based on Stockwell transform is proposed for intracranial long-term EEG data. The Stockwell transform is employed to obtain the time-frequency representation of the EEG signals, and then the power spectral density is calculated in the time-frequency plane to characterize the behavior of EEG recordings. After that, a classifier based on gradient boosting algorithm is used to make the classification. Finally, the postprocessing is utilized on the outputs of the classifier to obtain more stable and accurate detection results, which includes Kalman filter, threshold judgment, and collar technique. The performance of this method is assessed on the publicly available EEG database which contains approximately 533h of intracranial EEG recordings. The experimental results indicate that the proposed method can achieve a satisfactory sensitivity of 94.26%, a specificity of 96.34%, as well as a very short delay time of 0.56s.

Patient-specific ECG beat classification technique

Electrocardiogram (ECG) beat classification plays an important role in the timely diagnosis of the critical heart condition. An automated diagnostic system is proposed to classify five types of ECG classes, namely normal (N), ventricular ectopic beat (V), supra ventricular ectopic beat (S), fusion (F) and unknown (Q) as recommended by the Association for the Advancement of Medical Instrumentation (AAMI). The proposed method integrates the Stockwell transform (ST), a bacteria foraging optimisation (BFO) algorithm and a least mean square (LMS)-based multiclass support vector machine (SVM) classifier. The ST is utilised to extract the important morphological features which are concatenated with four timing features. The resultant combined feature vector is optimised by removing the redundant and irrelevant features using the BFO algorithm. The optimised feature vector is applied to the LMS-based multiclass SVM classifier for automated diagnosis. In the proposed technique, the LMS algorithm is used to modify the Lagrange multiplier, which in turn modifies the weight vector to minimise the classification error. The updated weights are used during the testing phase to classify ECG beats. The classification performances are evaluated using the MIT-BIH arrhythmia database. Average accuracy and sensitivity performances of the proposed system for V detection are 98.6% and 91.7%, respectively, and for S detections, 98.2% and 74.7%, respectively over the entire database. To generalise the capability, the classification performance is also evaluated using the St. Petersburg Institute of Cardiological Technics (INCART) database. The proposed technique performs better than other reported heartbeat techniques, with results suggesting better generalisation capability.