Wavelet transforms and the ECG: a review

An automatic sustained attention prediction (ASAP) method for infants and toddlers using wearable device signals

Sustained attention (SA) is a critical cognitive ability that emerges in infancy and affects various aspects of development. Research on SA typically occurs in lab settings, which may not reflect infants' real-world experiences. Infant wearable technology can collect multimodal data in natural environments, including physiological signals for measuring SA. Here we introduce an automatic sustained attention prediction (ASAP) method that harnesses electrocardiogram (ECG) and accelerometer (Acc) signals. Data from 75 infants (6- to 36-months) were recorded during different activities, with some activities emulating those occurring in the natural environment (i.e., free play). Human coders annotated the ECG data for SA periods validated by fixation data. ASAP was trained on temporal and spectral features from the ECG and Acc signals to detect SA, performing consistently across age groups. To demonstrate ASAP's applicability, we investigated the relationship between SA and perceptual features-saliency and clutter-measured from egocentric free-play videos. Results showed that saliency in infants' and toddlers' views increased during attention periods and decreased with age for attention but not inattention. We observed no differences between ASAP attention detection and human-coded SA periods, demonstrating that ASAP effectively detects SA in infants during free play. Coupled with wearable sensors, ASAP provides unprecedented opportunities for studying infant development in real-world settings.

Adaptive wavelet base selection for deep learning-based ECG diagnosis: A reinforcement learning approach

Electrocardiogram (ECG) signals are crucial in diagnosing cardiovascular diseases (CVDs). While wavelet-based feature extraction has demonstrated effectiveness in deep learning (DL)-based ECG diagnosis, selecting the optimal wavelet base poses a significant challenge, as it directly influences feature quality and diagnostic accuracy. Traditional methods typically rely on fixed wavelet bases chosen heuristically or through trial-and-error, which can fail to cover the distinct characteristics of individual ECG signals, leading to suboptimal performance. To address this limitation, we propose a reinforcement learning-based wavelet base selection (RLWBS) framework that dynamically customizes the wavelet base for each ECG signal. In this framework, a reinforcement learning (RL) agent iteratively optimizes its wavelet base selection (WBS) strategy based on successive feedback of classification performance, aiming to achieve progressively optimized feature extraction. Experiments conducted on the clinically collected PTB-XL dataset for ECG abnormality classification show that the proposed RLWBS framework could obtain more detailed time-frequency representation of ECG signals, yielding enhanced diagnostic performance compared to traditional WBS approaches.

Central venous pressure as a method of optimising atrio-ventricular delay after cardiac surgery

INTRODUCTION

Haemodynamic atrioventricular delay (AVD) optimisation has primarily focussed on signals that are not easy to acquire from a pacing system itself, such as invasive left ventricular catheterisation or arterial blood pressure (ABP). In this study, standard clinical central venous pressure (CVP) signals are tested as a potential alternative.

METHODS

Sixteen patients with a temporary pacemaker after cardiac surgery were studied. AV delay optimisation was performed by alternating between a reference AVD of 120ms and tested settings ranging from 40 to 280ms, with 8 replicates for each setting. Alongside (a) the raw data, three methods of correcting for respiration were tested: (b) limiting analysis to a respiratory cycle, (c) asymmetric least squares (ALS) and (d) discrete wavelet transform (DWT). The utility of a quality control step was tested.

RESULTS

CVP signals were a mirror image of the systolic ABP signals: The four R values were -0.674, -0.692, -0.631, -0.671 respectively (all p<0.001). With quality control, the mirror image was best for DWT (R = -0.76, p<0.001), with the CVP and ABP optima agreeing well (R = 0.78, p<0.001). The automated quality control signal correctly predicted the gap between the AVD optima calculated from ABP and CVP (R = 0.8, p<0.001).

CONCLUSIONS

Central venous pressure signals could be used to optimise AVD, because they have a reliable inverse relationship with ABP when pacemaker settings undergo protocolised testing. However, protocols need careful design to circumvent spontaneous biological variability.

Intraoperative circulation predict prolonged length of stay after head and neck free flap reconstruction: a retrospective study based on machine learning

Background

Head and neck free flap reconstruction presents challenges in managing intraoperative circulation, potentially leading to prolonged length of stay (PLOS). Limited research exists on the associations between intraoperative circulation and PLOS given the difficulty of manual quantification of intraoperative circulation time-series data. Therefore, this study aimed to quantify intraoperative circulation data and investigate its association with PLOS after free flap reconstruction utilizing machine learning algorithms.

Methods

804 patients who underwent head and neck free flap reconstruction between September 2019 and February 2021 were included. Machine learning tools (Fourier transform, et al.) were utilized to extract features to quantify intraoperative circulation data. To compare the accuracy of quantified intraoperative circulation and manual intraoperative circulation assessments in the PLOS prediction, predictive models based on these 2 assessment methods were developed and validated.

Results

Intraoperative circulation was quantified and a total of 114 features were extracted from intraoperative circulation data. Quantified intraoperative circulation models with a real-time predictive manner were constructed. A higher area under the receiver operating characteristic curve (AUROC) was observed in quantified intraoperative circulation data models (0.801 [95% CI, 0.733-0.869]) compared to manual intraoperative circulation assessment models (0.719 [95% CI, 0.641-0.797]) in PLOS prediction.

Conclusion

Machine learning algorithms facilitated quantification of intraoperative circulation data. The developed real-time quantified intraoperative circulation prediction models based on this quantification offer a potential strategy to optimize intraoperative circulation management and mitigate PLOS following head and neck free flap reconstruction.

Monitoring Age-Related Changes in Gait Complexity in the Wild with a Smartphone Accelerometer System

Stride-to-stride fluctuations during walking reflect age-related changes in gait adaptability and are estimated with nonlinear measures that confine data collection to controlled settings. Smartphones, with their embedded accelerometers, may provide accessible gait analysis throughout the day. This study investigated age-related differences in linear and nonlinear gait measures estimated from a smartphone accelerometer (SPAcc) in an unconstrained, free-living environment. Thirteen young adults (YA) and 11 older adults (OA) walked within a shopping mall with a SPAcc placed in their front right pants pocket. The inter-stride interval, calculated as the time difference between ipsilateral heel contacts, was used for dependent measures calculations. One-way repeated-measures analysis of variance revealed significant (p < 0.05) age-related differences (mean: YA, OA) for stride-time standard deviation (0.04 s, 0.05 s) and coefficient of variation (3.47%, 4.16%), sample entropy (SaEn) scale 1 (1.70, 1.86) and scale 3 (2.12, 1.80), and statistical persistence decay (31 strides, 23 strides). The fractal scaling index was not different between groups (0.93, 0.95), but exceeded those typically found in controlled settings, suggesting an upregulation in adaptive behaviour likely to accommodate the increased challenge of free-living walking. These findings support the SPAcc as a viable telehealth instrument for remote monitoring of gait dynamics, with implications for unsupervised fall-risk assessment.

Advanced Noise-Resistant Electrocardiography Classification Using Hybrid Wavelet-Median Denoising and a Convolutional Neural Network

The classification of ECG signals is a critical process because it guides the diagnosis of the proper treatment process for the patient. However, any form of disturbance with ECG signals can be highly conspicuous because of the mechanics involved in data acquisition from living beings, which has a significant impact on the classification procedure. The purpose of this research work is to advance ECG signal classification results by employing numerous denoising methods and, in turn, boost the accuracy of cardiovascular diagnoses. To simulate realistic conditions, we added various types of noise to ECG data, including Gaussian, salt and pepper, speckle, uniform, and exponential noise. To overcome the interference of noise from environments in the obtained ECG signals, we employed wavelet transform, median filter, Gaussian filter, and the hybrid of the wavelet and median filters. The proposed hybrid denoising method has better results than the other methods because of the use of wavelet multi-scale analysis and the ability of the median filter to avoid the loss of vital ECG characteristics. Thus, despite a certain proximity in the values, the hybrid method is significantly more accurate and reliable, as evidenced by the mean squared error (MSE), mean absolute error (MAE), R-squared, and Pearson correlation coefficient. More specifically, the hybrid approach provided an MSE of 0.0012 and an MAE of 0.025, the R-squared value for this study was 0.98, and the Pearson correlation coefficient was 0.99, which provides a very good resemblance to the original ECG confirmation. The proposed classification model is based on the modified lightweight CNN or MLCNN that was trained using the noisy and the denoised data. The findings demonstrated that by applying the denoised data, the testing accuracy, precision, recall, and F1 scores achieved 0.92, 0.91, 0.90, and 0.91 for the datasets, while the noisy data achieved 0.80, 0.78, 0.82, and 0.80, respectively. In this study, the signal quality and denoising methods were found to enhance ECG signal classification and diagnostic accuracy while encouraging proper preprocessing in future studies and applications for real-time ECG for cardiac care.

Motion artifact variability in biomagnetic wearable devices

Motion artifacts can be a significant noise source in biomagnetic measurements when magnetic sensors are not separated from the signal source. In ambient environments, motion artifacts can be up to ten times stronger than the desired signals, varying with environmental conditions. This study evaluates the variability of these artifacts and the effectiveness of a gradiometer in reducing them in such settings. To achieve these objectives, we first measured the single channel output in varying magnetic field conditions to observe the effect of homogeneous and gradient background fields. Our analysis revealed that the variability in motion artifact within an ambient environment is primarily influenced by the gradient magnetic field rather than the homogeneous one. Subsequently, we configured a gradiometer in parallel and vertical alignment with the direction of vibration (X-axis). Our findings indicated that in a gradient background magnetic field ranging from 1 nT/mm to 10 nT/mm, the single-channel sensor output exhibited a change of 164.97 pT per mm unit increase, while the gradiometer output showed a change of only 0.75 pT/mm within the same range. Upon repositioning the gradiometer vertically (Y direction), perpendicular to the direction of vibration, the single-channel output slope increased to 196.85 pT, whereas the gradiometer output only increased by 1.06 pT/mm for the same range. Our findings highlight the influence of ambient environments on motion artifacts and demonstrate the potential of gradiometers to mitigate these effects. In the future, we plan to record biomagnetic signals both inside and outside the shielded room to compare the efficacy of different gradiometer designs under varying environmental conditions.

An ECG denoising technique based on AHIN block and gradient difference max loss

The electrocardiogram (ECG) signal is susceptible to interference from unknown noises during the acquisition process due to their low frequency and amplitude, resulting in the loss of significant information in the signals. Recent advancements in deep learning models have shown promising results in signal processing. However, these models lack robustness against various types of noise and often overlook the gradient difference between denoised and original signals. In this study, we propose a novel deep learning denoising method based on an attention half instance normalization block (AHIN block) and a gradient difference max loss function (GDM Loss). Our approach consists of two stages: firstly, we input the noisy ECG signal to obtain a denoised version; secondly, we reconstruct the denoised signal by fusing preliminary results from the first stage while correcting waveform distortions caused by initial denoising to minimize information loss. Additionally, we introduce a new loss function that considers differences between slopes of the denoised ECG signal and clean ECG signal. To validate our proposed method's effectiveness, extensive experiments were conducted on both our model architecture and loss function compared with other state-of-the-art methods. Results demonstrate that our approach achieves excellent performance in terms of both signal-to-noise ratio (SNR) and root-mean-square error (RMSE). The proposed noise reduction method improves 8.86%, 12.05% and 10.50% respectively under BW, MA and EM noise.

Novel interpretable Feature set extraction and classification for accurate atrial fibrillation detection from ECGs

OBJECTIVE

We present a novel method for detecting atrial fibrillation (AFib) by analyzing Lead II electrocardiograms (ECGs) using a unique set of features.

METHODS

For this purpose, we used specific signal processing techniques, such as proper orthogonal decomposition, continuous wavelet transforms, discrete cosine transform, and standard cross-correlation, to extract 48 features from the ECGs. Thus, our approach aims to more effectively capture AFib signatures, such as beat-to-beat variability and fibrillatory waves, than traditional metrics. Moreover, our features were designed to be physiologically interpretable. Subsequently, we incorporated an XGBoost-based ECG classifier to train and evaluate it using the publicly available 'Training' dataset of the 2017 PhysioNet Challenge, which includes 'Normal,' 'AFib,' 'Other,' and 'Noisy' ECG categories.

RESULTS

Our method achieved an accuracy of 96 % and an F1-score of 0.83 for 'AFib' detection and 80 % accuracy and 0.85 F1-score for 'Normal' ECGs. Finally, we compared our method's performance with the top-classifiers from the 2017 PhysioNet Challenge, namely ENCASE, Random Forest, and Cascaded Binary. As reported by Clifford et al., 2017, these three best performing models scored 0.80, 0.83, 0.82, in terms of F1-score for 'AFib' detection, respectively.

CONCLUSION

Despite using significantly fewer features than the competition's state-of-the-art ECG detection algorithms (48 vs. 150-622), our model achieved a comparable F1-score of 0.83 for successful 'AFib' detection.

SIGNIFICANCE

The interpretable features specifically designed for 'AFib' detection results in our method's adaptability in clinical settings for real-time arrhythmia detection and is even effective with short ECGs (<10 heartbeats).

Miniaturized implantable temperature sensors for the long-term monitoring of chronic intestinal inflammation

Diagnosing and monitoring inflammatory bowel diseases, such as Crohn's disease, involves the use of endoscopic imaging, biopsies and serology. These infrequent tests cannot, however, identify sudden onsets and severe flare-ups to facilitate early intervention. Hence, about 70% of patients with Crohn's disease require surgical intestinal resections in their lifetime. Here we report wireless, miniaturized and implantable temperature sensors for the real-time chronic monitoring of disease progression, which we tested for nearly 4 months in a mouse model of Crohn's-disease-like ileitis. Local measurements of intestinal temperature via intraperitoneally implanted sensors held in place against abdominal muscular tissue via two sutures showed the development of ultradian rhythms at approximately 5 weeks before the visual emergence of inflammatory skip lesions. The ultradian rhythms showed correlations with variations in the concentrations of stress hormones and inflammatory cytokines in blood. Decreasing average temperatures over the span of approximately 23 weeks were accompanied by an increasing percentage of inflammatory species in ileal lesions. These miniaturized temperature sensors may aid the early treatment of inflammatory bowel diseases upon the detection of episodic flare-ups.

Uncertainty quantification in neural-network based pain intensity estimation

Improper pain management leads to severe physical or mental consequences, including suffering, a negative impact on quality of life, and an increased risk of opioid dependency. Assessing the presence and severity of pain is imperative to prevent such outcomes and determine the appropriate intervention. However, the evaluation of pain intensity is a challenging task because different individuals experience pain differently. To overcome this, many researchers in the field have employed machine learning models to evaluate pain intensity objectively using physiological signals. However, these efforts have primarily focused on pain point estimation, disregarding inherent uncertainty and variability in the data and model. A point estimate, which provides only partial information, is not sufficient for sound clinical decision-making. This study proposes a neural network-based method for objective pain interval estimation, and quantification of uncertainty. Our approach, which enables objective pain intensity estimation with desired confidence probabilities, affords clinicians a better understanding of a person's pain intensity. We explored three distinct algorithms: the bootstrap method, lower and upper bound estimation (LossL) optimized by genetic algorithm, and modified lower and upper bound estimation (LossS) optimized by gradient descent algorithm. Our empirical results demonstrate that LossS outperforms the other two by providing narrower prediction intervals. For 50%, 75%, 85%, and 95% prediction interval coverage probability, LossS provides average interval widths that are 22.4%, 7.9%, 16.7%, and 9.1% narrower than those of LossL, and 19.3%, 21.1%, 23.6%, and 26.9% narrower than those of bootstrap. As LossS outperforms, we assessed its performance in three different model-building approaches: (1) a generalized approach using a single model for the entire population, (2) a personalized approach with separate models for each individual, and (3) a hybrid approach with models for clusters of individuals. Results demonstrate that the hybrid model-building approach provides the best performance.

Energy wavelet signal processed ECG and standard 12 lead ECG: Diagnosis of early diastolic dysfunction

BACKGROUND

Left ventricular (LV) diastolic dysfunction (LVDD) is the result of impaired LV relaxation and identifies those at risk of developing heart failure. Echocardiography has been used as the gold standard to identify early LVDD. The signal processed electrocardiogram (hsECG) has demonstrated effectiveness to detect early LVDD. Whether or not the standard 12‑lead electrocardiogram (ECG) can accurately predict early LVDD is not known.

METHODS

A standard 12‑lead ECG including signal processing (hsECG) was performed in 569 patients. Patients with atrial fibrillation, bundle branch block, pre-excitation, left ventricular hypertrophy or known cardiovascular disease were excluded, leaving 464 examinations for analysis. Early LVDD was diagnosed by established methods using echocardiography. Repolarization abnormalities (T wave discordance) in V1, V6, I and aVL and the hsECG were compared to the echocardiographic findings to establish diagnostic accuracy.

RESULTS

A total of 84 (18.1%) patients were diagnosed with early LVDD. A combination of a borderline or abnormal finding on the hsECG produced the best diagnostic model (sensitivity 84.5%, specificity 47.9%). The best performing ECG lead was V1 with a sensitivity of 38.1% and specificity of 92.1%. Regression analysis demonstrated increasing age and V1 to be predictive of LVDD.

CONCLUSIONS

The hsECG displayed reasonable ability to detect early LVDD. Other than V1, repolarization abnormalities on the standard 12‑lead ECG did not. While lead V1 showed promise in detecting LVDD, whether this or any other simple ECG variable can predict future LVDD would be of further interest.

Evaluation of the responsiveness pattern to caffeine through a smart data-driven ECG non-linear multi-band analysis

This study aimed to explore more efficient ways of administering caffeine to the body by investigating the impact of caffeine on the modulation of the nervous system's activity through the analysis of electrocardiographic signals (ECG). An ECG non-linear multi-band analysis using Discrete Wavelet Transform (DWT) was employed to extract various features from healthy individuals exposed to different caffeine consumption methods: expresso coffee (EC), decaffeinated coffee (ED), Caffeine Oral Films (OF_caffeine), and placebo OF (OF_placebo). Non-linear feature distributions representing every ECG minute time series have been selected by PCA with different variance percentages to serve as inputs for 23 machine learning models in a leave-one-out cross-validation process for analyzing the behavior differences between ED/EC and OF_placebo/OF_caffeine groups, respectively, over time. The study generated 50-point accuracy curves per model, representing the discrimination power between groups throughout the 50 min. The best model accuracies for ED/EC varied between 30 and 70 %, (using the decision tree classifier) and OF_placebo/OF_caffeine ranged from 62 to 84 % (using Fine Gaussian). Notably, caffeine delivery through OFs demonstrated effective capacity compared to its placebo counterpart, as evidenced by significant differences in accuracy curves between OF_placebo/OF_caffeine. Caffeine delivery via OFs also exhibited rapid dissolution efficiency and controlled release rate over time, distinguishing it from EC. The study supports the potential of caffeine delivery through Caffeine OFs as a superior technology compared to traditional methods by means of ECG analysis. It highlights the efficiency of OFs in controlling the release of caffeine and underscores their promise for future caffeine delivery systems.

Observations of enhanced rainfall variability in Kenya, East Africa

Understanding local patterns of rainfall variability is of great concern in East Africa, where agricultural productivity is dominantly rainfall dependent. However, East African rainfall climatology is influenced by numerous drivers operating at multiple scales, and local patterns of variability are not adequately understood. Here, we show evidence of substantial variability of local rainfall patterns between 1981 and 2021 at the national and county level in Kenya, East Africa. Results show anomalous patterns of both wetting and drying in both the long and short rainy seasons, with evidence of increased frequency of extreme wet and dry events through time. Observations also indicate that seasonal and intraseasonal variability increased significantly after 2013, coincident with diminished coherence between ENSO (El Nino Southern Oscillation) and rainfall. Increasing frequency and magnitude of rainfall variability suggests increasing need for local-level climate change adaptation strategies.

Association of beat-to-beat P-wave analysis index to the extent of left atrial low-voltage areas in patients with paroxysmal atrial fibrillation

BACKGROUND

Left atrial (LA) fibrosis has been shown to be associated with atrial fibrillation (AF) recurrence. Beat-to-beat (B2B) index is a non-invasive classifier, based on B2B P-wave morphological and wavelet analysis, shown to be associated with AF incidence and recurrence. In this study, we tested the hypothesis that the B2B index is associated with the extent of LA low-voltage areas (LVAs) on electroanatomical mapping.

METHODS

Patients with paroxysmal AF scheduled for pulmonary vein isolation, without evident structural remodeling, were included. Pre-ablation electroanatomical voltage maps were used to calculate the surface of LVAs (<0.5 mV). B2B index was compared between patients with small versus large LVAs.

RESULTS

35 patients were included (87% male, median age 62). The median surface area of LVAs was 7.7 (4.4-15.8) cm2 corresponding to 5.6 (3.3-12.1) % of LA endocardial surface. B2B index was 0.57 (0.52-0.59) in patients with small LVAs (below the median) compared to 0.65 (0.56-0.77) in those with large LVAs (above the median) (p = 0.009). In the receiver operator characteristic curve analysis for predicting large LVAs, the c-statistic was 0.75 (p = 0.006) for B2B index and 0.81 for the multivariable model including B2B index (multivariable p = 0.04) and P-wave duration.

CONCLUSION

In patients with paroxysmal AF without overt atrial myopathy, B2B P-wave analysis appears to be a useful non-invasive correlate of low-voltage areas-and thus fibrosis-in the LA. This finding establishes a pathophysiological basis for B2B index and its potential usefulness in the selection process of patients who are likely to benefit most from further invasive treatment.

Detection of preceding sleep apnea using ECG spectrogram during CPAP titration night: A novel machine-learning and bag-of-features framework

Obstructive sleep apnea (OSA) has a heavy health-related burden on patients and the healthcare system. Continuous positive airway pressure (CPAP) is effective in treating OSA, but adherence to it is often inadequate. A promising solution is to detect sleep apnea events in advance, and to adjust the pressure accordingly, which could improve the long-term use of CPAP treatment. The use of CPAP titration data may reflect a similar response of patients to therapy at home. Our study aimed to develop a machine-learning algorithm using retrospective electrocardiogram (ECG) data and CPAP titration to forecast sleep apnea events before they happen. We employed a support vector machine (SVM), k-nearest neighbour (KNN), decision tree (DT), and linear discriminative analysis (LDA) to detect sleep apnea events 30-90 s in advance. Preprocessed 30 s segments were time-frequency transformed to spectrograms using continuous wavelet transform, followed by feature generation using the bag-of-features technique. Specific frequency bands of 0.5-50 Hz, 0.8-10 Hz, and 8-50 Hz were also extracted to detect the most detected band. Our results indicated that SVM outperformed KNN, LDA, and DT across frequency bands and leading time segments. The 8-50 Hz frequency band gave the best accuracy of 98.2%, and a F1-score of 0.93. Segments 60 s before sleep events seemed to exhibit better performance than other pre-OSA segments. Our findings demonstrate the feasibility of detecting sleep apnea events in advance using only a single-lead ECG signal at CPAP titration, making our proposed framework a novel and promising approach to managing obstructive sleep apnea at home.

Deep Representation Learning With Sample Generation and Augmented Attention Module for Imbalanced ECG Classification

Developing an efficient heartbeat monitoring system has become a focal point in numerous healthcare applications. Specifically, in the last few years, heartbeat classification for arrhythmia detection has gained considerable interest from researchers. This paper presents a novel deep representation learning method for the efficient detection of arrhythmic beats. To mitigate the issues associated with the imbalanced data distribution, a novel re-sampling strategy is introduced. Unlike the existing oversampling methods, the proposed technique transforms majority-class samples into minority-class samples with a novel translation loss function. This approach assists the model in learning a more generalized representation of crucially important minority class samples. Moreover, by exploiting an auxiliary feature, an augmented attention module is designed that focuses on the most relevant and target-specific information. We adopted an inter-patient classification paradigm to evaluate the proposed method. The experimental results of this study on the MIT-BIH arrhythmia database clearly indicate that the proposed model with augmented attention mechanism and over-sampling strategy significantly learns a balanced deep representation and improves the classification performance of vital heartbeats.

Electrical Impedance Tomography as a monitoring tool during weaning from mechanical ventilation: an observational study during the spontaneous breathing trial

BACKGROUND

Prolonged weaning from mechanical ventilation is associated with poor clinical outcome. Therefore, choosing the right moment for weaning and extubation is essential. Electrical Impedance Tomography (EIT) is a promising innovative lung monitoring technique, but its role in supporting weaning decisions is yet uncertain. We aimed to evaluate physiological trends during a T-piece spontaneous breathing trail (SBT) as measured with EIT and the relation between EIT parameters and SBT success or failure.

METHODS

This is an observational study in which twenty-four adult patients receiving mechanical ventilation performed an SBT. EIT monitoring was performed around the SBT. Multiple EIT parameters including the end-expiratory lung impedance (EELI), delta Tidal Impedance (ΔZ), Global Inhomogeneity index (GI), Rapid Shallow Breathing Index (RSBIEIT), Respiratory Rate (RREIT) and Minute Ventilation (MVEIT) were computed on a breath-by-breath basis from stable tidal breathing periods.

RESULTS

EELI values dropped after the start of the SBT (p < 0.001) and did not recover to baseline after restarting mechanical ventilation. The ΔZ dropped (p < 0.001) but restored to baseline within seconds after restarting mechanical ventilation. Five patients failed the SBT, the GI (p = 0.01) and transcutaneous CO2 (p < 0.001) values significantly increased during the SBT in patients who failed the SBT compared to patients with a successful SBT.

CONCLUSION

EIT has the potential to assess changes in ventilation distribution and quantify the inhomogeneity of the lungs during the SBT. High lung inhomogeneity was found during SBT failure. Insight into physiological trends for the individual patient can be obtained with EIT during weaning from mechanical ventilation, but its role in predicting weaning failure requires further study.

Automated identification of atrial fibrillation from single-lead ECGs using multi-branching ResNet

Introduction: Atrial fibrillation (AF) is the most common cardiac arrhythmia, which is clinically identified with irregular and rapid heartbeat rhythm. AF puts a patient at risk of forming blood clots, which can eventually lead to heart failure, stroke, or even sudden death. Electrocardiography (ECG), which involves acquiring bioelectrical signals from the body surface to reflect heart activity, is a standard procedure for detecting AF. However, the occurrence of AF is often intermittent, costing a significant amount of time and effort from medical doctors to identify AF episodes. Moreover, human error is inevitable, as even experienced medical professionals can overlook or misinterpret subtle signs of AF. As such, it is of critical importance to develop an advanced analytical model that can automatically interpret ECG signals and provide decision support for AF diagnostics. Methods: In this paper, we propose an innovative deep-learning method for automated AF identification using single-lead ECGs. We first extract time-frequency features from ECG signals using continuous wavelet transform (CWT). Second, the convolutional neural networks enhanced with residual learning (ReNet) are employed as the functional approximator to interpret the time-frequency features extracted by CWT. Third, we propose to incorporate a multi-branching structure into the ResNet to address the issue of class imbalance, where normal ECGs significantly outnumber instances of AF in ECG datasets. Results and Discussion: We evaluate the proposed Multi-branching Resnet with CWT (CWT-MB-Resnet) with two ECG datasets, i.e., PhysioNet/CinC challenge 2017 and ECGs obtained from the University of Oklahoma Health Sciences Center (OUHSC). The proposed CWT-MB-Resnet demonstrates robust prediction performance, achieving an F1 score of 0.8865 for the PhysioNet dataset and 0.7369 for the OUHSC dataset. The experimental results signify the model's superior capability in balancing precision and recall, which is a desired attribute for ensuring reliable medical diagnoses.

CLINet: A novel deep learning network for ECG signal classification

Machine learning is poised to revolutionize medicine with algorithms that spot cardiac arrhythmia. An automated diagnostic approach can boost the efficacy of diagnosing life-threatening arrhythmia disorders in routine medical procedures. In this paper, we propose a deep learning network CLINet for ECG signal classification. Our network uses convolution, LSTM and involution layers to bring their unique advantages together. For both convolution and involution layers, we use multiple, large size kernels for multi-scale representation learning. CLINet does not require complicated pre-processing and can handle electrocardiograms of any length. Our network achieves 99.90% accuracy on the ICCAD dataset and 99.94% accuracy on the MIT-BIH dataset. With only 297 K parameters, our model can be easily embedded in smart wearable devices. The source code of CLINet is available at https://github.com/CandleLabAI/CLINet-ECG-Classification-2024.

The Impact of Feature Extraction on Classification Accuracy Examined by Employing a Signal Transformer to Classify Hand Gestures Using Surface Electromyography Signals

Interest in developing techniques for acquiring and decoding biological signals is on the rise in the research community. This interest spans various applications, with a particular focus on prosthetic control and rehabilitation, where achieving precise hand gesture recognition using surface electromyography signals is crucial due to the complexity and variability of surface electromyography data. Advanced signal processing and data analysis techniques are required to effectively extract meaningful information from these signals. In our study, we utilized three datasets: NinaPro Database 1, CapgMyo Database A, and CapgMyo Database B. These datasets were chosen for their open-source availability and established role in evaluating surface electromyography classifiers. Hand gesture recognition using surface electromyography signals draws inspiration from image classification algorithms, leading to the introduction and development of the Novel Signal Transformer. We systematically investigated two feature extraction techniques for surface electromyography signals: the Fast Fourier Transform and wavelet-based feature extraction. Our study demonstrated significant advancements in surface electromyography signal classification, particularly in the Ninapro database 1 and CapgMyo dataset A, surpassing existing results in the literature. The newly introduced Signal Transformer outperformed traditional Convolutional Neural Networks by excelling in capturing structural details and incorporating global information from image-like signals through robust basis functions. Additionally, the inclusion of an attention mechanism within the Signal Transformer highlighted the significance of electrode readings, improving classification accuracy. These findings underscore the potential of the Signal Transformer as a powerful tool for precise and effective surface electromyography signal classification, promising applications in prosthetic control and rehabilitation.

Robust algorithm for the detection and classification of QRS complexes with different morphologies using the continuous spline wavelet transform with automatic scale detection

This work presents an algorithm for the detection and classification of QRS complexes based on the continuous wavelet transform (CWT) with splines. This approach can evaluate the CWT at any integer scale and the analysis is not restricted to powers of two. The QRS detector comprises four stages: implementation of CWT with splines, detection of QRS complexes, searching for undetected QRS complexes, and correction of the R wave peak location in detected QRS complexes. After, the onsets and ends of the QRS complexes are detected. The algorithm was evaluated with synthetic ECG and with the manually annotated databases: MIT-BIH Arrhythmia, European ST-T, QT and PTB Diagnostic ECG. Evaluation results of the QRS detector were: MIT-BIH arrhythmia database (109,447 beats analyzed), sensitivity Se = 99.72% and positive predictivity P+ = 99.87%; European ST-T database (790522 beats analyzed), Se = 99.92% and P+ = 99.55% and QT database (86498 beats analyzed), Se = 99.97% and P+ = 99.99%. To evaluate the delineation algorithm of the QRS onset (Qi) and QRS end (J) with the QT and PTB Diagnostic ECG databases, the mean and standard deviations of the differences between the automatic and manual annotated location of these points were calculated. The standard deviations were close to the accepted tolerances for deviations determined by the CSE experts. The proposed algorithm is robust to noise, artifacts and baseline drifts, classifies QRS complexes, automatically selects the CWT scale according to the sampling frequency of the ECG record used, and adapts to changes in the heart rate, amplitude and morphology of QRS complexes.

Representative QRS loop of the VCG record evaluation

Introduction: This study proposes an algorithm for preprocessing VCG records to obtain a representative QRS loop. Methods: The proposed algorithm uses the following methods: Digital filtering to remove noise from the signal, wavelet-based detection of ECG fiducial points and isoelectric PQ intervals, spatial alignment of QRS loops, QRS time synchronization using root mean square error minimization and ectopic QRS elimination. The representative QRS loop is calculated as the average of all QRS loops in the VCG record. The algorithm is evaluated on 161 VCG records from a database of 58 healthy control subjects, 69 patients with myocardial infarction, and 34 patients with bundle branch block. The morphologic intra-individual beat-to-beat variability rate is calculated for each VCG record. Results and Discussion: The maximum relative deviation is 12.2% for healthy control subjects, 19.3% for patients with myocardial infarction, and 17.2% for patients with bundle branch block. The performance of the algorithm is assessed by measuring the morphologic variability before and after QRS time synchronization and ectopic QRS elimination. The variability is reduced by a factor of 0.36 for healthy control subjects, 0.38 for patients with myocardial infarction, and 0.41 for patients with bundle branch block. The proposed algorithm can be used to generate a representative QRS loop for each VCG record. This representative QRS loop can be used to visualize, compare, and further process VCG records for automatic VCG record classification.

Discrete Wavelet Transformation for the Sensitive Detection of Ultrashort Radiation Pulse with Radiation-induced Acoustics

Radiation-induced acoustics (RIA) shows promise in advancing radiological imaging and radiotherapy dosimetry methods. However, RIA signals often require extensive averaging to achieve reasonable signal-to-noise ratios, which increases patient radiation exposure and limits real-time applications. Therefore, this paper proposes a discrete wavelet transform (DWT) based filtering approach to denoise the RIA signals and avoid extensive averaging. The algorithm was benchmarked against low-pass filters and tested on various types of RIA sources, including low-energy X-rays, high-energy X-rays, and protons. The proposed method significantly reduced the required averages (1000 times less averaging for low-energy X-ray RIA, 32 times less averaging for high-energy X-ray RIA, and 4 times less averaging for proton RIA) and demonstrated robustness in filtering signals from different sources of radiation. The coif5 wavelet in conjunction with the sqtwolog threshold selection algorithm yielded the best results. The proposed DWT filtering method enables high-quality, automated, and robust filtering of RIA signals, with a performance similar to low-pass filtering, aiding in the clinical translation of radiation-based acoustic imaging for radiology and radiation oncology.

Novel Entropy-Based Metrics for Long-Term Atrial Fibrillation Recurrence Prediction Following Surgical Ablation: Insights from Preoperative Electrocardiographic Analysis

Atrial fibrillation (AF) is a prevalent cardiac arrhythmia often treated concomitantly with other cardiac interventions through the Cox-Maze procedure. This highly invasive intervention is still linked to a long-term recurrence rate of approximately 35% in permanent AF patients. The aim of this study is to preoperatively predict long-term AF recurrence post-surgery through the analysis of atrial activity (AA) organization from non-invasive electrocardiographic (ECG) recordings. A dataset comprising ECGs from 53 patients with permanent AF who had undergone Cox-Maze concomitant surgery was analyzed. The AA was extracted from the lead V1 of these recordings and then characterized using novel predictors, such as the mean and standard deviation of the relative wavelet energy (RWEm and RWEs) across different scales, and an entropy-based metric that computes the stationary wavelet entropy variability (SWEnV). The individual predictors exhibited limited predictive capabilities to anticipate the outcome of the procedure, with the SWEnV yielding a classification accuracy (Acc) of 68.07%. However, the assessment of the RWEs for the seventh scale (RWEs7), which encompassed frequencies associated with the AA, stood out as the most promising individual predictor, with sensitivity (Se) and specificity (Sp) values of 80.83% and 67.09%, respectively, and an Acc of almost 75%. Diverse multivariate decision tree-based models were constructed for prediction, giving priority to simplicity in the interpretation of the forecasting methodology. In fact, the combination of the SWEnV and RWEs7 consistently outperformed the individual predictors and excelled in predicting post-surgery outcomes one year after the Cox-Maze procedure, with Se, Sp, and Acc values of approximately 80%, thus surpassing the results of previous studies based on anatomical predictors associated with atrial function or clinical data. These findings emphasize the crucial role of preoperative patient-specific ECG signal analysis in tailoring post-surgical care, enhancing clinical decision making, and improving long-term clinical outcomes.

BRAVEHEART: Open-source software for automated electrocardiographic and vectorcardiographic analysis

BACKGROUND AND OBJECTIVES

Electrocardiographic (ECG) and vectorcardiographic (VCG) analyses are used to diagnose current cardiovascular disease and for risk stratification for future adverse cardiovascular events. With increasing use of digital ECGs, research into novel ECG/VCG parameters has increased, but widespread computer-based ECG/VCG analysis is limited because there are no currently available, open-source, and easily customizable software packages designed for automated and reproducible analysis.

METHODS AND RESULTS

We present BRAVEHEART, an open-source, modular, customizable, and easy to use software package implemented in the MATLAB programming language, for scientific analysis of standard 12-lead ECGs acquired in a digital format. BRAVEHEART accepts a wide variety of digital ECG formats and provides complete and automatic ECG/VCG processing with signal denoising to remove high- and low-frequency artifact, non-dominant beat identification and removal, accurate fiducial point annotation, VCG construction, median beat construction, customizable measurements on median beats, and output of measurements and results in numeric and graphical formats.

CONCLUSIONS

The BRAVEHEART software package provides easily customizable scientific analysis of ECGs and VCGs. We hope that making BRAVEHART available will allow other researchers to further the field of ECG/VCG analysis without having to spend significant time and resources developing their own ECG/VCG analysis software and will improve the reproducibility of future studies. Source code, compiled executables, and a detailed user guide can be found at http://github.com/BIVectors/BRAVEHEART. The source code is distributed under the GNU General Public License version 3.

Implementation of Wavelet-Transform-Based Algorithms in an FPGA for Heart Rate and RT Interval Automatic Measurements in Real Time: Application in a Long-Term Ambulatory Electrocardiogram Monitor

Cardiovascular diseases are currently the leading cause of death worldwide. Thus, there is a need for non-invasive ambulatory (Holter) ECG monitors with automatic measurements of ECG intervals to evaluate electrocardiographic abnormalities of patients with cardiac diseases. This work presents the implementation of algorithms in an FPGA for beat-to-beat heart rate and RT interval measurements based on the continuous wavelet transform (CWT) with splines for a prototype of an ambulatory ECG monitor of three leads. The prototype's main elements are an analog-digital converter ADS1294, an FPGA of Xilinx XC7A35T-ICPG236C of the Artix-7 family of low consumption, immersed in a low-scale Cmod-A7 development card integration, an LCD display and a micro-SD memory of 16 Gb. A main state machine initializes and manages the simultaneous acquisition of three leads from the ADS1294 and filters the signals using a FIR filter. The algorithm based on the CWT with splines detects the QRS complex (R or S wave) and then the T-wave end using a search window. Finally, the heart rate (60/RR interval) and the RT interval (from R peak to T-wave end) are calculated for analysis of its dynamics. The micro-SD memory stores the three leads and the RR and RT intervals, and an LCD screen displays the beat-to-beat values of heart rate, RT interval and the electrode connection. The algorithm implemented on the FPGA achieved satisfactory results in detecting different morphologies of QRS complexes and T wave in real time for the analysis of heart rate and RT interval dynamics.

MRI radiomics-based evaluation of tuberculous and brucella spondylitis

OBJECTIVES

We analyzed magnetic resonance imaging (MRI) and radiomics labels from tuberculous spondylitis (TBS) and brucella spondylitis (BS) to build machine learning models that differentiate TBS from BS and culture-positive TBS (TBS(+)) from culture-negative TBS (TBS(-).

METHODS

This retrospective study included 56 patients with BS, 63 patients with TBS(+) and 71 patients with TBS(-). Radiomics labels were extracted from T2-weighted fat-suppression images. MRI labels were analyzed via logistic regression (LR); radiomics labels were analyzed by t-tests, SelectKBest, and least absolute shrinkage and selection operator (LASSO). Random forest (RF) and support vector machine (SVM) models were established using radiomics or joint (radiomics+MRI) labels. Models were evaluated by receiver operating characteristic curves, areas under the curve (AUCs), decision curve analysis (DCA), and Hosmer-Lemeshow tests.

RESULTS

When joint-label models were used to compare BS vs TBS(+) and BS vs TBS(-) groups, SVM AUCs were 0.904 and 0.944, respectively, whereas RF AUCs were 0.950 and 0.947, respectively; these were higher than the AUCs of the MRI label-based LR model. DCA showed that radiomics-based machine learning models had a greater net benefit; Hosmer-Lemeshow tests demonstrated good prediction consistency for all models.

CONCLUSIONS

Radiomics can help distinguish TBS from BS and TBS(+) from TBS(-).

DWT-CNNTRN: a Convolutional Transformer for ECG Classification with Discrete Wavelet Transform

Cardiovascular diseases are the leading cause of death worldwide. The diagnoses of cardiovascular diseases are usually carried out by cardiologists utilizing Electrocardiograms (ECGs). To assist these physicians in making an accurate diagnosis, there is a growing need for reliable and automatic ECG classifiers.In this study, a new method is proposed to classify 12-lead ECG recordings. The proposed model is composed of four components: the CNN(Convolutional Neural Network) module, the transformer module, the global hybrid pooling layer, and a classification layer. To improve the classification performance, the model takes the discrete wavelet transform of ECG signals as the model inputs and utilizes a hybrid pooling layer to condense the most important features over each period.The proposed model is evaluated using the test set of the China Physiological Signal Challenge 2018 dataset with 12-lead ECGs. It performs with an average accuracy of 0.86 and an average F1-scores of 0.83. The scores are particularly good for the block conditions (LBBB, RBBB, I-AVB). The main advantage of the proposed model is that, it obtains good results with a significantly smaller number of parameters compared to other individual and ensemble models.Clinical relevance- This work establishes a new ECG classifier model with high performance and low model size. It can make automatic ECG analysis more accessible, efficient, and accurate, especially in remote or underserved areas.

A movable unshielded magnetocardiography system

Magnetocardiography (MCG), which uses high-sensitivity magnetometers to record magnetic field signals generated by electrical activity in the heart, is a noninvasive method for evaluating heart diseases such as arrhythmia and ischemia. The MCG measurements usually require the participant keeping still in a magnetically shielded room due to the immovable sensor and noisy external environments. These requirements limit MCG applications, such as exercise MCG tests and long-term MCG observations, which are useful for early detections of heart diseases. Here, we introduce a movable MCG system that can clearly record MCG signals of freely behaving participants in an unshielded environment. On the basis of optically pumped magnetometers with a sensitivity of 140 fT/Hz1/2, we successfully demonstrated the resting MCG and the exercise MCG tests. Our method is promising to realize a practical movable multichannel unshielded MCG system that nearly sets no limits to participants and brings another kind of insight into the medical diagnosis of heart disease.

Hierarchical deep learning with Generative Adversarial Network for automatic cardiac diagnosis from ECG signals

Cardiac disease is the leading cause of death in the US. Accurate heart disease detection is critical to timely medical treatment to save patients' lives. Routine use of the electrocardiogram (ECG) is the most common method for physicians to assess the cardiac electrical activities and detect possible abnormal conditions. Fully utilizing the ECG data for reliable heart disease detection depends on developing effective analytical models. In this paper, we propose a two-level hierarchical deep learning framework with Generative Adversarial Network (GAN) for ECG signal analysis. The first-level model is composed of a Memory-Augmented Deep AutoEncoder with GAN (MadeGAN), which aims to differentiate abnormal signals from normal ECGs for anomaly detection. The second-level learning aims at robust multi-class classification for different arrhythmia identification, which is achieved by integrating the transfer learning technique to transfer knowledge from the first-level learning with the multi-branching architecture to handle the data-lacking and imbalanced data issues. We evaluate the performance of the proposed framework using real-world ECG data from the MIT-BIH arrhythmia database. Experimental results show that our proposed model outperforms existing methods that are commonly used in current practice.

A dynamic learning-based ECG feature extraction method for myocardial infarction detection

Objective.Myocardial infarction (MI) is one of the leading causes of human mortality in all cardiovascular diseases globally. Currently, the 12-lead electrocardiogram (ECG) is widely used as a first-line diagnostic tool for MI. However, visual inspection of pathological ECG variations induced by MI remains a great challenge for cardiologists, since pathological changes are usually complex and slight.Approach.To have an accuracy of the MI detection, the prominent features extracted from in-depth mining of ECG signals need to be explored. In this study, a dynamic learning algorithm is applied to discover prominent features for identifying MI patients via mining the hidden inherent dynamics in ECG signals. Firstly, the distinctive dynamic features extracted from the multi-scale decomposition of dynamic modeling of the ECG signals effectively and comprehensibly represent the pathological ECG changes. Secondly, a few most important dynamic features are filtered through a hybrid feature selection algorithm based on filter and wrapper to form a representative reduced feature set. Finally, different classifiers based on the reduced feature set are trained and tested on the public PTB dataset and an independent clinical data set.Main results.Our proposed method achieves a significant improvement in detecting MI patients under the inter-patient paradigm, with an accuracy of 94.75%, sensitivity of 94.18%, and specificity of 96.33% on the PTB dataset. Furthermore, classifiers trained on PTB are verified on the test data set collected from 200 patients, yielding a maximum accuracy of 84.96%, sensitivity of 85.04%, and specificity of 84.80%.Significance.The experimental results demonstrate that our method performs distinctive dynamic feature extraction and may be used as an effective auxiliary tool to diagnose MI patients.

Identification of Coronary Culprit Lesion in ST Elevation Myocardial Infarction by Using Deep Learning

OBJECTIVE

Early revascularization of the occluded coronary artery in patients with ST elevation myocardial infarction (STEMI) has been demonstrated to decrease mortality and morbidity. Currently, physicians rely on features of electrocardiograms (ECGs) to identify the most likely location of coronary arteries related to an infarct. We sought to predict these culprit arteries more accurately by using deep learning.

METHODS

A deep learning model with a convolutional neural network (CNN) that incorporated ECG signals was trained on 384 patients with STEMI who underwent primary percutaneous coronary intervention (PCI) at a medical center. The performances of various signal preprocessing methods (short-time Fourier transform [STFT] and continuous wavelet transform [CWT]) with different lengths of input ECG signals were compared. The sensitivity and specificity for predicting each infarct-related artery and the overall accuracy were evaluated.

RESULTS

ECG signal preprocessing with STFT achieved fair overall prediction accuracy (79.3%). The sensitivity and specificity for predicting the left anterior descending artery (LAD) as the culprit vessel were 85.7% and 88.4%, respectively. The sensitivity and specificity for predicting the left circumflex artery (LCX) were 37% and 99%, respectively, and the sensitivity and specificity for predicting the right coronary artery (RCA) were 88.4% and 82.4%, respectively. Using CWT (Morlet wavelet) for signal preprocessing resulted in better overall accuracy (83.7%) compared with STFT preprocessing. The sensitivity and specificity were 93.46% and 80.39% for LAD, 56% and 99.7% for LCX, and 85.9% and 92.9% for RCA, respectively.

CONCLUSION

Our study demonstrated that deep learning with a CNN could facilitate the identification of the culprit coronary artery in patients with STEMI. Preprocessing ECG signals with CWT was demonstrated to be superior to doing so with STFT.

Study on exercise muscle fatigue based on sEMG and ECG data fusion and temporal convolutional network

BACKGROUND

Muscle fatigue is a crucial indicator to determine whether training is in place and to protect trainers.

PURPOSE

To make full use of morphological information of surface EMG and ECG signals in the time domain, a new idea and method for the fatigue assessment of exercise muscles based on data fusion is proposed in this paper.

METHODS

sEMG and ECG time series with the same length were obtained by signal preprocessing and sequence normalization, feature extraction of sequence tenses was realized by a deep learning network based on sequential convolution and signal fusion model of muscle fatigue evaluation was established by D-S evidence theory.

EXPERIMENT

Thirty volunteers were recruited and divided into three groups. ECG signals and sEMG signals at the biceps brachii of the right upper limb were monitored in a 20-minute exercise cycle.

RESULTS

The prediction result of TCN based on time domain signal is better than the commonly used KNN and SVM recognition algorithm, and the recognition accuracy of relaxed, excessive and fatigue by D-S fusion was 89%, 86%, 88.5%. The accuracy was 0.9055, 0.9494 and 0.9269, respectively. The recall rates of the three conditions were 0.9303, 0.9570 and 0.9435. The F-score of the three conditions was 0.8911, 0.8764 and 0.8837, respectively.

CONCLUSION

Based on time series and time series convolutional network, sEMG and ECG fusion of motor muscle recognition method can better distinguish different state information and has certain practical value in the fields of muscle evaluation, clinical diagnosis, wearable devices and so on.

Armband Sensors Location Assessment for Left Arm-ECG Bipolar Leads Waveform Components Discovery Tendencies around the MUAC Line

Sudden cardiac death (SCD) risk can be reduced by early detection of short-lived and transient cardiac arrhythmias using long-term electrocardiographic (ECG) monitoring. Early detection of ventricular arrhythmias can reduce the risk of SCD by allowing appropriate interventions. Long-term continuous ECG monitoring, using a non-invasive armband-based wearable device is an appealing solution for detecting early heart rhythm abnormalities. However, there is a paucity of understanding on the number and best bipolar ECG electrode pairs axial orientation around the left mid-upper arm circumference (MUAC) for such devices. This study addresses the question on the best axial orientation of ECG bipolar electrode pairs around the left MUAC in non-invasive armband-based wearable devices, for the early detection of heart rhythm abnormalities. A total of 18 subjects with almost same BMI values in the WASTCArD arm-ECG database were selected to assess arm-ECG bipolar leads quality using proposed metrics of relative (normalized) signal strength measurement, arm-ECG detection performance of the main ECG waveform event component (QRS) and heart-rate variability (HRV) in six derived bipolar arm ECG-lead sensor pairs around the armband circumference, having regularly spaced axis angles (at 30° steps) orientation. The analysis revealed that the angular range from -30° to +30°of arm-lead sensors pair axis orientation around the arm, including the 0° axis (which is co-planar to chest plane), provided the best orientation on the arm for reasonably good QRS detection; presenting the highest sensitivity (Se) median value of 93.3%, precision PPV median value at 99.6%; HRV RMS correlation (p) of 0.97 and coefficient of determination (R²) of 0.95 with HRV gold standard values measured in the standard Lead-I ECG.

A novel deep learning model to detect COVID-19 based on wavelet features extracted from Mel-scale spectrogram of patients' cough and breathing sounds

The goal of this paper is to classify the various cough and breath sounds of COVID-19 artefacts in the signals from dynamic real-life environments. The main reason for choosing cough and breath sounds than other common symptoms to detect COVID-19 patients from the comfort of their homes, so that they do not overload the Medicare system and therefore do not unwittingly spread the disease by regularly monitoring themselves. The presented model includes two main phases. The first phase is the sound-to-image transformation, which is improved by the Mel-scale spectrogram approach. The second phase consists of extraction of features and classification using nine deep transfer models (ResNet18/34/50/100/101, GoogLeNet, SqueezeNet, MobileNetv2, and NasNetmobile). The dataset contains information data from almost 1600 people (1185 Male and 415 Female) from all over the world. Our classification model is the most accurate, its accuracy is 99.2% according to the SGDM optimizer. The accuracy is good enough that a large set of labelled cough and breath data may be used to check the possibility for generalization. The results demonstrate that ResNet18 is the best stable model for classifying cough and breath tones from a restricted dataset, with a sensitivity of 98.3% and a specificity of 97.8%. Finally, the presented model is shown to be more trustworthy and accurate than any other present model. Cough and breath study accuracy is promising enough to put extrapolation and generalization to the test.

Experimental Investigation on the Bioprotective Role of Trehalose on Glutamine Solutions by Infrared Spectroscopy

Glutamine plays a significant role in several basic metabolic processes and is an important regulator of heat shock protein response. The present work is focused on the analysis of the thermal response of aqueous solutions of Glutamine and aqueous solutions of Glutamine in the presence of Trehalose by means of infrared absorption technique. The performed study shows how in the case of a multicomponent system, characterized by a huge number of spectral contributions whose assignment are questionable, the Spectral Distance (SD) and the Cross Wavelet Correlation (XWT) approaches are able to furnish explanatory parameters that can characterize the variations in the spectra behaviour, which is an efficient tool for quantitative comparisons. With this purpose, the analysis has been performed by evaluating the SD and the XWT parameters for the whole investigated spectral range, i.e., 4000–400 cm⁻¹, for scans collected as a function of temperature in the range 20 °C ÷ 60 °C both for Glutamine/Water compounds and for Glutamine /Water/Trehalose mixtures. By means of these analyses, it is found that in aqueous solutions of Glutamine, with respect to aqueous solutions of Glutamine in the presence of Trehalose, the SD and XWT temperature trends follow a linear behaviour where the angular coefficient for Glutamine /Water/Trehalose compounds are lower than that of the Glutamine-Water system in both cases. The obtained findings suggest that Trehalose stabilizes Glutamine against heat treatment.

Discrete Wavelet Transform Analysis of the Electroretinogram in Autism Spectrum Disorder and Attention Deficit Hyperactivity Disorder

Background

To evaluate the electroretinogram waveform in autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD) using a discrete wavelet transform (DWT) approach.

Methods

A total of 55 ASD, 15 ADHD and 156 control individuals took part in this study. Full field light-adapted electroretinograms (ERGs) were recorded using a Troland protocol, accounting for pupil size, with five flash strengths ranging from –0.12 to 1.20 log photopic cd.s.m^–2. A DWT analysis was performed using the Haar wavelet on the waveforms to examine the energy within the time windows of the a- and b-waves and the oscillatory potentials (OPs) which yielded six DWT coefficients related to these parameters. The central frequency bands were from 20–160 Hz relating to the a-wave, b-wave and OPs represented by the coefficients: a20, a40, b20, b40, op80, and op160, respectively. In addition, the b-wave amplitude and percentage energy contribution of the OPs (%OPs) in the total ERG broadband energy was evaluated.

Results

There were significant group differences (p < 0.001) in the coefficients corresponding to energies in the b-wave (b20, b40) and OPs (op80 and op160) as well as the b-wave amplitude. Notable differences between the ADHD and control groups were found in the b20 and b40 coefficients. In contrast, the greatest differences between the ASD and control group were found in the op80 and op160 coefficients. The b-wave amplitude showed both ASD and ADHD significant group differences from the control participants, for flash strengths greater than 0.4 log photopic cd.s.m^–2 (p < 0.001).

Conclusion

This methodological approach may provide insights about neuronal activity in studies investigating group differences where retinal signaling may be altered through neurodevelopment or neurodegenerative conditions. However, further work will be required to determine if retinal signal analysis can offer a classification model for neurodevelopmental conditions in which there is a co-occurrence such as ASD and ADHD.

Reliable Detection of Myocardial Ischemia Using Machine Learning Based on Temporal-Spatial Characteristics of Electrocardiogram and Vectorcardiogram

Background: Myocardial ischemia is a common early symptom of cardiovascular disease (CVD). Reliable detection of myocardial ischemia using computer-aided analysis of electrocardiograms (ECG) provides an important reference for early diagnosis of CVD. The vectorcardiogram (VCG) could improve the performance of ECG-based myocardial ischemia detection by affording temporal-spatial characteristics related to myocardial ischemia and capturing subtle changes in ST-T segment in continuous cardiac cycles. We aim to investigate if the combination of ECG and VCG could improve the performance of machine learning algorithms in automatic myocardial ischemia detection. Methods: The ST-T segments of 20-second, 12-lead ECGs, and VCGs were extracted from 377 patients with myocardial ischemia and 52 healthy controls. Then, sample entropy (SampEn, of 12 ECG leads and of three VCG leads), spatial heterogeneity index (SHI, of VCG) and temporal heterogeneity index (THI, of VCG) are calculated. Using a grid search, four SampEn and two features are selected as input signal features for ECG-only and VCG-only models based on support vector machine (SVM), respectively. Similarly, three features (S I , THI, and SHI, where S I is the SampEn of lead I) are further selected for the ECG + VCG model. 5-fold cross validation was used to assess the performance of ECG-only, VCG-only, and ECG + VCG models. To fully evaluate the algorithmic generalization ability, the model with the best performance was selected and tested on a third independent dataset of 148 patients with myocardial ischemia and 52 healthy controls. Results: The ECG + VCG model with three features (S I ,THI, and SHI) yields better classifying results than ECG-only and VCG-only models with the average accuracy of 0.903, sensitivity of 0.903, specificity of 0.905, F1 score of 0.942, and AUC of 0.904, which shows better performance with fewer features compared with existing works. On the third independent dataset, the testing showed an AUC of 0.814. Conclusion: The SVM algorithm based on the ECG + VCG model could reliably detect myocardial ischemia, providing a potential tool to assist cardiologists in the early diagnosis of CVD in routine screening during primary care services.

Coarse-graining and the Haar wavelet transform for multiscale analysis

BACKGROUND

Multiscale entropy (MSE) has become increasingly common as a quantitative tool for analysis of physiological signals. The MSE computation involves first decomposing a signal into multiple sub-signal 'scales' using a coarse-graining algorithm.

METHODS

The coarse-graining algorithm averages adjacent values in a time series to produce a coarser scale time series. The Haar wavelet transform convolutes a time series with a scaled square wave function to produce an approximation which is equivalent to averaging points.

RESULTS

Coarse-graining is mathematically identical to the Haar wavelet transform approximations. Thus, multiscale entropy is entropy computed on sub-signals derived from approximations of the Haar wavelet transform. By describing coarse-graining algorithms properly as Haar wavelet transforms, the meaning of 'scales' as wavelet approximations becomes transparent. The computed value of entropy is different with different wavelet basis functions, suggesting further research is needed to determine optimal methods for computing multiscale entropy.

CONCLUSION

Coarse-graining is mathematically identical to Haar wavelet approximations at power-of-two scales. Referring to coarse-graining as a Haar wavelet transform motivates research into the optimal approach to signal decomposition for entropy analysis.

Recent Advances in Materials and Flexible Sensors for Arrhythmia Detection

Arrhythmias are one of the leading causes of death in the United States, and their early detection is essential for patient wellness. However, traditional arrhythmia diagnosis by expert evaluation from intermittent clinical examinations is time-consuming and often lacks quantitative data. Modern wearable sensors and machine learning algorithms have attempted to alleviate this problem by providing continuous monitoring and real-time arrhythmia detection. However, current devices are still largely limited by the fundamental mismatch between skin and sensor, giving way to motion artifacts. Additionally, the desirable qualities of flexibility, robustness, breathability, adhesiveness, stretchability, and durability cannot all be met at once. Flexible sensors have improved upon the current clinical arrhythmia detection methods by following the topography of skin and reducing the natural interface mismatch between cardiac monitoring sensors and human skin. Flexible bioelectric, optoelectronic, ultrasonic, and mechanoelectrical sensors have been demonstrated to provide essential information about heart-rate variability, which is crucial in detecting and classifying arrhythmias. In this review, we analyze the current trends in flexible wearable sensors for cardiac monitoring and the efficacy of these devices for arrhythmia detection.

Combining Measures of Signal Complexity and Machine Learning for Time Series Analyis: A Review

Measures of signal complexity, such as the Hurst exponent, the fractal dimension, and the Spectrum of Lyapunov exponents, are used in time series analysis to give estimates on persistency, anti-persistency, fluctuations and predictability of the data under study. They have proven beneficial when doing time series prediction using machine and deep learning and tell what features may be relevant for predicting time-series and establishing complexity features. Further, the performance of machine learning approaches can be improved, taking into account the complexity of the data under study, e.g., adapting the employed algorithm to the inherent long-term memory of the data. In this article, we provide a review of complexity and entropy measures in combination with machine learning approaches. We give a comprehensive review of relevant publications, suggesting the use of fractal or complexity-measure concepts to improve existing machine or deep learning approaches. Additionally, we evaluate applications of these concepts and examine if they can be helpful in predicting and analyzing time series using machine and deep learning. Finally, we give a list of a total of six ways to combine machine learning and measures of signal complexity as found in the literature.

Advanced e-Call Support Based on Non-Intrusive Driver Condition Monitoring for Connected and Autonomous Vehicles

BACKGROUND

The growth of the number of vehicles in traffic has led to an exponential increase in the number of road accidents with many negative consequences, such as loss of lives and pollution.

METHODS

This article focuses on using a new technology in automotive electronics by equipping a semi-autonomous vehicle with a complex sensor structure that is able to provide centralized information regarding the physiological signals (Electro encephalogram-EEG, electrocardiogram-ECG) of the driver/passengers and their location along with indoor temperature changes, employing the Internet of Things (IoT) technology. Thus, transforming the vehicle into a mobile sensor connected to the internet will help highlight and create a new perspective on the cognitive and physiological conditions of passengers, which is useful for specific applications, such as health management and a more effective intervention in case of road accidents. These sensor structures mounted in vehicles will allow for a higher detection rate of potential dangers in real time. The approach uses detection, recording, and transmission of relevant health information in the event of an incident as support for e-Call or other emergency services, including telemedicine.

RESULTS

The novelty of the research is based on the design of specialized non-invasive sensors for the acquisition of EEG and ECG signals installed in the headrest and backrest of car seats, on the algorithms used for data analysis and fusion, but also on the implementation of an IoT temperature measurement system in several points that simultaneously uses sensors based on MEMS technology. The solution can also be integrated with an e-Call system for telemedicine emergency assistance.

CONCLUSION

The research presents both positive and negative results of field experiments, with possible further developments. In this context, the solution has been developed based on state-of-the-art technical devices, methods, and technologies for monitoring vital functions of the driver/passengers (degree of fatigue, cognitive state, heart rate, blood pressure). The purpose is to reduce the risk of accidents for semi-autonomous vehicles and to also monitor the condition of passengers in the case of autonomous vehicles for providing first aid in a timely manner. Reported abnormal values of vital parameters (critical situations) will allow interveneing in a timely manner, saving the patient's life, with the support of the e-Call system.

Electromyography Signal Analysis and Classification using Time-Frequency Representations and Deep Learning

Analysis and classification of electromyography (EMG) signals are crucial for rehabilitation and motor control. This study investigates electromyogram (EMG) time-frequency representations and then creates conventional and deep learning models for EMG signal classification. Firstly, a dataset of single-channel surface EMG signals has been recorded for four subjects to differentiate between forearm flexion and extension. Then, different time-frequency EMG representations have been used to build conventional and deep learning models for EMG classification. We compared the performance of pre-trained convolutional neural network models, namely GoogLeNet, SqueezeNet and AlexNet, and achieved accuracies of 92.71%, 90.63% and 87.5%, respectively. Also, data augmentation techniques on the levels of raw EMG signals and their time- frequency representations helped improve the accuracy of GoogLeNet to 96.88%. Furthermore, our approach demonstrated superior performance on another publicly available 10-class EMG dataset, and also using traditional classifiers trained on hand-crafted features.

Sleep Phases in Crayfish: Relationship Between Brain Electrical Activity and Autonomic Variables

In vertebrates like mammals and birds, two types of sleep have been identified: rapid eye movement and non-rapid eye movement sleep. Each one is associated with specific electroencephalogram patterns and is accompanied by variations in cardiac and respiratory frequencies. Sleep has been demonstrated only in a handful of invertebrates, and evidence for different sleep stages remains elusive. Previous results show that crayfish sleeps while lying on one side on the surface of the water, but it is not known if this animal has sleep phases. Heart rate and respiratory frequency are modified by diverse changes in the crayfish environment during wakefulness, and previously, we showed that variations in these variables are present during sleep despite that there are no autonomic anatomical structures described in this animal. Here, we conducted experiments to search for sleep phases in crayfish and the relationships between sleep and cardiorespiratory activity. We used the wavelet transform, grouping analysis with k-means clustering, and principal component analysis, to analyze brain and cardiorespiratory electrical activity. Our results show that (a) crayfish can sleep lying on one side or when it is motionless and (b) the depth of sleep (measured as the power of electroencephalographic activity) changes over time and is accompanied by oscillations in cardiorespiratory signal amplitude and power. Finally, we propose that in crayfish there are at least three phases of sleep.

Wavelet Analysis and Self-Similarity of Photoplethysmography Signals for HRV Estimation and Quality Assessment

Peak-to-peak intervals in Photoplethysmography (PPG) can be used for heart rate variability (HRV) estimation if the PPG is collected from a healthy person at rest. Many factors, such as a person's movements or hardware issues, can affect the signal quality and make some parts of the PPG signal unsuitable for reliable peak detection. Therefore, a robust HRV estimation algorithm should not only detect peaks, but also identify corrupted signal parts. We introduce such an algorithm in this paper. It uses continuous wavelet transform (CWT) for peak detection and a combination of features derived from CWT and metrics based on PPG signals' self-similarity to identify corrupted parts. We tested the algorithm on three different datasets: a newly introduced Welltory-PPG-dataset containing PPG signals collected with smartphones using the Welltory app, and two publicly available PPG datasets: TROIKAand PPG-DaLiA. The algorithm demonstrated good accuracy in peak-to-peak intervals detection and HRV metric estimation.