Electrocardiogram Signal Data Research Articles

Analysing and Evaluating the Performance of Deep-Learning-Based Arrhythmia Detection Using Electrocardiogram Signals

Cardiac arrhythmia is a cardiac irregularity that impacts a significant number of individuals globally. Certain arrhythmias may be benign or occur only once, while recurring arrhythmias have the potential to cause organ failure, increase the risk of stroke by a factor of five, and even lead to sudden cardiac death. Thus, in order to identify and treat arrhythmia and prevent potentially fatal cardiac problems, the rapid detection and categorization of Electrocardiogram (ECG) signals is of paramount importance. The non-invasive technique employs electrodes to examine the electrical potentials of the heart, facilitating the detection of structural and functional irregularities that contribute to the diagnosis of cardiovascular illnesses. Nevertheless, the high likelihood of manual interpretation, which is both time-consuming and susceptible to error caused by weariness, poses a significant challenge for cardiologists in identifying and diagnosing cardiac issues. In recent times, there has been a growing utilization of Deep Learning (DL) models in the field of arrhythmia prediction, with the aim of enhancing clinical decision-making and potentially mitigating the likelihood of valetudinarian fatality. These models enhance the diagnostic capabilities of electrocardiography by detecting pathological situations, extracting anatomically meaningful data, evaluating cardio-motion, and assessing the quality of echo images. As a result, they serve as an alternative tool for precise diagnosis and treatment of arrhythmias. In order to forecast and categorize arrhythmias based on ECG signal data, this study provides a comprehensive review of various deep learning approaches. The initial section of the study provides a concise overview of various deep learning-based prediction models that have been developed by multiple researchers for echocardiography systems. Subsequently, a comparative analysis is undertaken to comprehend the limitations of those algorithms and propose a novel approach to improve the accuracy of cardiac view classification in echocardiography systems.

IoT-based health information system using MitApp for abnormal electrocardiogram signal monitoring

Information systems are currently developing very rapidly, and this is inseparable from the role of internet of things (IoT) technology, especially in the world of telemedicine. MitApp is an open-source application that can be used to monitor electrocardiogram (ECG) signals in real-time. The aim of this study is to develop an IoT-based ECG signal monitoring system that utilizes the MitApp application to detect abnormal ECG signals that are characterized by symptoms of cardiac arrhythmias. To process ECG signal data obtained from lead electrode results, the research method utilizes Arduino Uno as a microcontroller. The result is then displayed on the thin film transistor (TFT) layer using the Nextion module. The ESP32 module is used as a Wi-Fi module to send data to the MitApp app on a smartphone. The results showed that the results of the comparison test of ECG signal module data with ECG simulator tools with beats per minute values of 60, 80, 100, 120, and 140 obtained an error rate of 0.05. Based on these results, there is potential to develop this feature and integrate the system with the patient management system to improve the effectiveness of remote monitoring.

Arrhythmia classification for non-experts using infinite impulse response (IIR)-filter-based machine learning and deep learning models of the electrocardiogram.

Arrhythmias are a leading cause of cardiovascular morbidity and mortality. Portable electrocardiogram (ECG) monitors have been used for decades to monitor patients with arrhythmias. These monitors provide real-time data on cardiac activity to identify irregular heartbeats. However, rhythm monitoring and wave detection, especially in the 12-lead ECG, make it difficult to interpret the ECG analysis by correlating it with the condition of the patient. Moreover, even experienced practitioners find ECG analysis challenging. All of this is due to the noise in ECG readings and the frequencies at which the noise occurs. The primary objective of this research is to remove noise and extract features from ECG signals using the proposed infinite impulse response (IIR) filter to improve ECG quality, which can be better understood by non-experts. For this purpose, this study used ECG signal data from the Massachusetts Institute of Technology Beth Israel Hospital (MIT-BIH) database. This allows the acquired data to be easily evaluated using machine learning (ML) and deep learning (DL) models and classified as rhythms. To achieve accurate results, we applied hyperparameter (HP)-tuning for ML classifiers and fine-tuning (FT) for DL models. This study also examined the categorization of arrhythmias using different filters and the changes in accuracy. As a result, when all models were evaluated, DenseNet-121 without FT achieved 99% accuracy, while FT showed better results with 99.97% accuracy.

Effectiveness of Wavelet and Fourier Transform Methods for Denoising ECG Apnea Signals

An Electrocardiogram (ECG) signal results from recording the heart's electrical activity. Currently, diseases originating from heart abnormalities still dominate the world. Therefore, doctors need to find an ECG signal that is free from noise, this cleaning process is called denoising. The denoising methods commonly used are the wavelet transform method, and the Fourier transform method. Testing was conducted on ECG signal data with abnormalities in Apnea with RR intervals filled with noise, making the QRS complex difficult to identify. As a result, the Discrete Wavelet Transform (DWT) method produces the best form of reconstruction, resembling the initial ECG signal, while the Fast Fourier transform (FFT) method produces the best signal-to-noise ratio (SNR) for the denoising process. So, we get a superior FFT method in the ECG signal denoising process.

Explainable Deep Learning-Based Approach for Multilabel Classification of Electrocardiogram

Recently computer-aided diagnosis methods have been widely adopted to aid doctors in disease diagnosis making their decisions more reliable and error-free. Electrocardiogram (ECG) is the most commonly used, noninvasive diagnostic tool for investigating various cardiovascular diseases. In real life, patients suffer from more than one heart disease at a time. So any practical automated heart disease diagnosis system should identify multiple heart diseases present in a single ECG signal. In this article, we propose a novel deep learning-based method for the multilabel classification of ECG signals. The proposed method can accurately identify up to two labels of an ECG signal pertaining to eight rhythm or morphological abnormalities of the heart and also the normal heart condition. Also, the black-box nature of deep learning models prevents them from being applied to high-risk decisions like the automated heart disease diagnosis. So in this article, we also establish an explainable artificial intelligence (XAI) framework for ECG classification using class activation maps obtained from the Grad-CAM technique. In the proposed method, we train a convolutional neural network (CNN) with constructed ECG matrices. With the experiments conducted, we establish that training the CNN by taking only one label for each ECG signal data point is enough for the network to learn the features of an ECG point with multilabel information in it (multiple heart diseases at the same time). During classification, we apply thresholding on the output probabilities of the softmax layer of our CNN, to obtain the multilabel classification of ECG signals.We trained the model with 6311 ECG records and tested the model with 280 ECG records. During testing, the model achieved a subset accuracy of 96.2% and a hamming loss of 0.037 and a precision of 0.986 and a recall of 0.949 and an F1-score of 0.967. Considering the fact that the model has performed very well in all the metrics of multilabel classification, the model can be directly used as a practical tool for automated heart disease diagnosis.

Turning waste into wealth: Person identification by emotion‐disturbed electrocardiogram

AbstractThe issue of electrocardiogram (ECG)‐based person identification has attracted intense research interests nowadays. Different than existing related researches that advocate accentuating useful information and attenuating noisy artefacts in sensor data processing, A novel strategy of ‘turning waste into wealth’ is proposed to exploit the new discriminative information from the relationship between noise disturbance and signal data for this issue. Specifically, the authors design a new and simple method, the Set‐Group Distance Measure, based on the suitable fusion of multiple minority‐based distance measurements, whose power has initially been discovered for the issue. This method takes advantage of the collaborative variation information from the relative relationship, which is named as ‘relative information’, between different types of emotional noise disturbances and ECG signal data, to tackle the problem of large intra‐class variation but small inter‐class difference during identification. Experimental results have demonstrated the reasonability, effectiveness, robustness, efficiency and practicability of the proposed method upon public benchmark databases. This proposal not only provides technological inspirations for the further study in ECG‐based person identification, but also shows a fresh feasible way to handle the noise‐signal relationship for more general topics of sensor data classification.

Congestive heart failure detection based on attention mechanism-enabled bi-directional long short-term memory model in the internet of medical things

As congestive heart failure (CHF) has become an important medical problem facing the world, the detection of the CHF is very important. Internet of medical things (IoMT) has been increasing popularity as its application and function in industry 4.0 era. IoMT systems provide an opportunity for innovations and cross-discipline applications in the medical fields, that can significantly improve the efficiency. With the help of IoMT systems, patients can have in-home and on-body sensors that monitor their vitals easily and constantly, which can provide massive data for detecting disease. However, the most important thing is to find a good model to the automatic detection for massive medical data collected by IoMT systems. In this case, the paper proposes attention mechanism-enabled bi-directional long short-term memory (ABLSTM) model based on Electrocardiogram (ECG) signals, to automatically detect the CHF in IoMT systems. This method makes full use of the model features that can record long-term and short-term signal information, as well as the functions of attention mechanism with adaptive learning in local features, and effectively extracts complex features of ECG signals and performs detection. ECG signal data is from two public datasets, to train and test the proposed ABLSTM model. At the same time, for the cases with noise and data differences, we propose a preprocessing process for ECG signals, and discuss the impact of different data segmentation methods on the model performance. The experimental results show that the proposed ABLSTM model has the highest accuracy rate with 96.6% for the CHF detection, which is higher than other four baseline methods. Therefore, this proposed method can achieve a good result in the detection of the CHF.

A new intelligent ECG recognition approach based on CNN and improved ALO-SVM

Cardiovascular disease is one of the most common diseases, which seriously threatens people’s life and health. Therefore, cardiovascular disease prevention becomes one of the most attractive research topics in health care system design. Intelligent recognition of electrocardiogram (ECG) signals represents an effective method for rapid diagnosis and the evaluation of cardiovascular diseases in medicine. Realization and efficiency of the classification of ECG signals in real time play major roles in the detection of cardiovascular diseases. This paper is concerned with the proposition of an intelligent ECG signal recognition method based on a convolutional neural network (CNN) and support vector machines (SVM) with an improved antlion algorithm (ALO). First, the ECG signal is denoised and pre-processed by lifting the wavelet. Subsequently, CNN is used to extract the signal characteristics of the denoising signal, and the extracted signal characteristics are used as the input characteristics of the SVM. Finally, an improved ALO algorithm is used to optimize the relevant input functions of the SVM to achieve a better signal classification. In our algorithm, the performance is enhanced by improving the threshold estimation method of the lifting wavelet, to improve the filtering effect. The proposed CNN architecture is tested with multi-lead ECG signals from the MIT-BIH ECG signal data set. The results display that the method has obtained an average accuracy, sensitivity, and specificity values of \(99.97\%\), \(99.97\%\), and \(99.99\%\), respectively. Compared with the existing results, the proposed approach has a better recognition performance.

QRS Detection in Electrocardiogram Signal of Exercise Physical Activity

Accurate estimation of heart beats from electrocardiogram (ECG) signals during exercise activity is a very challenging problem. Unlike standard ECG, the signals recorded during exercise activity can affect the accuracy of QRS detection due to the noises and artifacts arise from body movements activity. However, most of the studies on QRS detection often used clean and standard data for the evaluations and assumed to reflect the overall performance of detectors. In addition, there are not many methods that evaluated using real ECG signal in their studies especially during the exercise activity. Therefore, this study was undertaken to access and evaluated the performance of QRS detection algorithms on the real ECG signal data. Three well-known QRS detection algorithms were re-implemented in this study. The ECG signal recorded under realistic movement conditions and can serving as a realistic data to assess the performance of QRS detection is used. The performance of the algorithm in real ECG signal data in sitting, walking, and jogging was then presented. The results show the algorithms capable to detect the QRS in ECG signal of exercise activity.

Data prediction for cases of incorrect data in multi-node electrocardiogram monitoring

The development of a mesh topology in multi-node electrocardiogram (ECG) monitoring based on the ZigBee protocol still has limitations. When more than one active ECG node sends a data stream, there will be incorrect data or damage due to a failure of synchronization. The incorrect data will affect signal interpretation. Therefore, a mechanism is needed to correct or predict the damaged data. In this study, the method of expectation-maximization (EM) and regression imputation (RI) was proposed to overcome these problems. Real data from previous studies are the main modalities used in this study. The ECG signal data that has been predicted is then compared with the actual ECG data stored in the main controller memory. Root mean square error (RMSE) is calculated to measure system performance. The simulation was performed on 13 ECG waves, each of them has 1000 samples. The simulation results show that the EM method has a lower predictive error value than the RI method. The average RMSE for the EM and RI methods is 4.77 and 6.63, respectively. The proposed method is expected to be used in the case of multi-node ECG monitoring, especially in the ZigBee application to minimize errors.

Premature Ventricular Contraction Recognition Based on a Deep Learning Approach

Electrocardiogram signal (ECG) is considered a significant biological signal employed to diagnose heart diseases. An ECG signal allows the demonstration of the cyclical contraction and relaxation of human heart muscles. This signal is a primary and noninvasive tool employed to recognize the actual life threat related to the heart. Abnormal ECG heartbeat and arrhythmia are the possible symptoms of severe heart diseases that can lead to death. Premature ventricular contraction (PVC) is one of the most common arrhythmias which begins from the lower chamber of the heart and can cause cardiac arrest, palpitation, and other symptoms affecting all activities of a patient. Nowadays, computer-assisted techniques reduce doctors' burden to assess heart arrhythmia and heart disease automatically. In this study, we propose a PVC recognition based on a deep learning approach using the MIT-BIH arrhythmia database. Firstly, 10 heartbeat and statistical features including three morphological features (RS amplitude, QR amplitude, and QRS width) and seven statistical features are computed for each signal. The extraction process of these features is conducted for 20 s of ECG data that create a feature vector. Next, these features are fed into a convolutional neural network (CNN) to find unique patterns and classify them more effectively. The obtained results prove that our pipeline improves the diagnosis performance more effectively.

Designing an Optimum and Reduced Order Filter for Efficient ECG QRS Peak Detection and Classification of Arrhythmia Data.

Electrocardiogram (ECG) is commonly used biological signals that show an important role in cardiac analysis. The interpretation and acquisition of QRS complex are significant measures of ECG data dispensation. The R wave has a vital character in the analysis of cardiac rhythm irregularities as well as in the determination of heart rate variability (HRV). This manuscript is proposed to design a new artificial-intelligence-based approach of QRS peak detection and classification of the ECG data. The design of reduced order IIR filter is proposed for the low pass smoothening of the ECG signal data. The min-max optimization is used for optimizing the filter coefficient to design the reduced order filter. In this research paper, elimination of baseline wondering and the power line interferences from the ECG signal is of main attention. The result presented that the accuracy is increased by around 13% over the basic Pan-Tompkins method and around 8% over the existing FIR-filter-based classification rules.

Design of Robust Heart Abnormality Detection System based on Wavelet Denoising Algorithm

The technology that continues to be developed by many researchers today is an automatic heart attack detection system based on an Electrocardiogram (ECG) signal. Several other studies have been carried out to build an Internet of Things (IoT) based heart abnormality detection system. Based on the analysis of related studies that have been carried out previously, several researchers have developed an ECG signal-based heart abnormality detection system using clean ECG signal data. While the reality of the concept of an IoT-based detection system, the process of recording ECG signal data on the sensor, the process of sending data to the server, and the process of retrieving data from the server are vulnerable to noise exposure. ECG signal containing noise will greatly affect the accuracy of system detection. This paper proposes the development of a noise-resistant heart condition detection system using a wavelet denoising algorithm. The process of denoising ECG signals using the Wavelet algorithm is generally able to improve the accuracy of detecting noisy ECG signals. The most significant increase in accuracy is seen in the low SNR value. The Daubechies 4 (db4) denoising algorithm is the best-performing algorithm. The ECG signal classification method uses the Artificial Neural Network (ANN) algorithm. Detection system hardware is also designed in this research using the concept based on the Internet of Things.

Low-Cost Multimodal Physiological Telemonitoring System through Internet of Things

The objective of this study is to develop and test a patient telemonitoring system. This study was encouraged by the high number of health workers fatalities in Indonesia due to physical contact without proper protection. Based on the symptoms of COVID-19 it consists of electrocardiogram (ECG) sensors, body temperature sensors, respiratory rate sensors, and pulse oximeter. The physiological data were captured by the sensors and collected by a microcontroller then it sends the data to a cloud system so that health workers can access the data. The experiments were performed to test both the offline and online protocol to compare data sent via a direct connection and data sent via Wi-Fi. In the offline testing, there were several limitations observed such as the low sampling frequency of the ECG signals that reduce the fidelity of the signals. Such problems were also observed on respiratory rate data. Furthermore, the system is also very prone to subjects’ movement-related noise. The measurements of peripheral oxygen saturation (SpO2) and body temperature, on the other hand, have been detected the slight change up to 0.1% and 0.5oC respectively. In the online testing, the data transmission to the cloud is sent per 30 seconds so that morphologically the ECG signal data are not representative. The system requires a lot of improvements and future study should be directed to improve signals acquisition and processing while maintaining the concept of low-cost. Design improvement should also include a better attachment design to the human body as well as greater data transmission for the online system.

Analysis of Differences in ECG Characteristics for Different Types of Drivers under Anxiety

Anxiety is a complex emotion characterized by an unpleasant feeling of tension when people anticipate a threat or negative consequence. It is regarded as a comprehensive reflection of human thought processes, physiological arousal, and external stimuli. The actual state of emotion can be represented objectively by human physiological signals. This study aims to analyze the differences of ECG (electrocardiogram) characteristics for various types of drivers under anxiety. We used several methods to induce drivers’ mood states (calm and anxiety) and then conducted the real and virtual driving experiments to collect driver’s ECG signal data. Physiological changes in ECG during the experiments were recorded using the PSYLAB software. The independent sample t‐test analysis was conducted to determine if there are significant differences in ECG characteristics for different types of drivers in anxious state during driving. The results show that there are significant differences in ECG signal characteristics of drivers by gender, age, and driving experience, in time domain, frequency domain, and waveform under anxiety. Our findings of this study contribute to the development of more intelligent and personalized driver warning system, which could improve road traffic safety.

ARTIFICIAL INTELLIGENCE-ENABLED ECG ALGORITHM FOR THE SCREENING OF DIABETES

We aimed to develop an artificial intelligence (AI) based electrocardiogram (ECG) algorithm to screen patients for DM. A convolutional neural network was used to identify diabetes using ECG signal data in the CARTeGENE registry (population-based study in Quebec, Canada). In the registry, 7,460

Predictors of hemodynamic decompensation in progressive hypovolemia: Compensatory reserve versus heart rate variability.

Hemorrhage remains the leading cause of death following traumatic injury in both civilian and military settings. Heart rate variability (HRV) and heart rate complexity (HRC) have been proposed as potential "new vital signs" for monitoring trauma patients; however, the added benefit of HRV or HRC for decision support remains unclear. Another new paradigm, the compensatory reserve measurement (CRM), represents the integration of all cardiopulmonary mechanisms responsible for compensation during relative blood loss and was developed to identify current physiologic status by estimating the progression toward hemodynamic decompensation. In the present study, we hypothesized that CRM would provide greater sensitivity and specificity to detect progressive reductions in central circulating blood volume and onset of decompensation as compared with measurements of HRV and HRC. Continuous, noninvasive measurements of compensatory reserve and electrocardiogram signals were made on 101 healthy volunteers during lower-body negative pressure (LBNP) to the point of decompensation. Measures of HRV and HRC were taken from electrocardiogram signal data. Compensatory reserve measurement demonstrated a superior sensitivity and specificity (receiver operator characteristic area under the curve [ROC AUC] = 0.93) compared with all HRV measures (ROC AUC ≤ 0.84) and all HRC measures (ROC AUC ≤ 0.86). Sensitivity and specificity values at the ROC optimal thresholds were greater for CRM (sensitivity = 0.84; specificity = 0.84) than HRV (sensitivity, ≤0.78; specificity, ≤0.77), and HRC (sensitivity, ≤0.79; specificity, ≤0.77). With standardized values across all levels of LBNP, CRM had a steeper decline, less variability, and explained a greater proportion of the variation in the data than both HRV and HRC during progressive hypovolemia. These findings add to the growing body of literature describing the advantages of CRM for detecting reductions in central blood volume. Most importantly, these results provide further support for the potential use of CRM in the triage and monitoring of patients at highest risk for the onset of shock following blood loss.

Measures of Compensatory Reserve are More Sensitive and Specific than Heart Rate Variability as Early Predictors of Hemodynamic Decompensation

BackgroundHemorrhage remains the leading cause of death following traumatic injury in both civilian and military settings. Heart rate variability (HRV) and complexity (HRC) have been recognized as potential “new vital signs” for monitoring trauma patients, although their benefit for triage decision support remains unclear. Another new paradigm, the compensatory reserve measurement (CRM), represents the integration of all mechanisms responsible for compensation during blood loss, designed to identify current physiologic status by estimating the progression toward hemodynamic decompensation. We hypothesized that CRM would provide greater sensitivity and specificity to detect progressive reductions in central circulating blood volume and onset of decompensation as compared to measurements of HRV and HRC.MethodsContinuous, noninvasive measurements of CRM and electrocardiogram (ECG) were made on 101 healthy volunteers (59 males, 42 females; mean±SD age 28±8 years) during stepwise progressive central hypovolemia to the point of decompensation. Lower body negative pressure (LBNP) was used to simulate hemorrhage by inducing the onset and progression of central hypovolemia. Time‐domain measures of HRV were taken from ECG signal data including R‐to‐R Interval Standard Deviation in milliseconds (ms), Root Mean Square Standard Deviation, percentage of RRI that vary by at least 50 ms, and amplitude measures of High Frequency Oscillations and Low Frequency Oscillations (CDMLF) obtained from the complex demodulation method. Measures of HRC included Sample Entropy, Fractal Dimension by Dispersion Analysis, Stationarity and Detrended Fluctuation Analysis. Data were analyzed using generalized estimating equations with compound symmetry covariance structures for longitudinal correlated data analysis of continuous variables, the results of which were used to report least‐square means with 95% confidence intervals at each LBNP level.ResultsCRM demonstrated superior ROC AUC (0.93) compared with all measures of HRV (≤ 0.84), and HRC (≤ 0.86). Sensitivity and specificity values at the ROC optimal thresholds were greater for CRM (SENS=0.84; SPEC=0.84) than HRV (SENS ≤ 0.78; SPEC ≤ 0.77) and HRC (SENS ≤ 0.79; SPEC ≤ 0.77). With standardized values across all levels of LBNP, CRM had a steeper decline, less variability, and explained a greater proportion of the variation in the data than both HRV and HRC during progressive hypovolemia.ConclusionConsistent with our hypothesis, CRM had greater sensitivity and specificity in detecting progressive reductions in circulating blood volume and decompensation than HRV and HRC. These findings add to the growing body of literature describing the advantages of CRM for detecting reductions in central blood volume. Most importantly, these results provide further support for the potential use of CRM in the triage and monitoring of patients at risk of circulatory shock following blood loss.Support or Funding InformationStudy funding was provided by a grant from the US Army Combat Casualty Care Research Program (D‐009‐2014‐USAISR).This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

The Real-time Electrocardiogram Signal Monitoring System in Wireless Sensor Network

The Cardiovascular disease (CVD) is the most of death in the world. Electrocardiogram (ECG) is the graph that shows heart electrical activities. The physician record and detect the abnormal Electrocardiogram (ECG) signal by the Holter monitor that patient need to carry on the device for record ECG signal in 24 hours. Pan-Tomkins algorithm was appropriate for Real-time ECG signal recognition because high accuracy and rapidly analysis. This research propose the Real-time ECG Signal monitoring system for detect the abnormal ECG signal by using Pan-Tomkins algorithm with Wireless Sensor Network. The system separated into 2 part; sender module and receiver module. Experimental the system by using the ECG signal data from MIT-BIH database. Selected 20 samples of abnormal ECG signal then experimental at 10 and 20 meters sender module-receiver module distance, calculate R-R interval and R amplitude threshold The results show that the Real-time ECG signal monitoring system detect 17 abnormal ECG signal, the accuracy is 85%. This systems efficient for detect the abnormal of ECG signal in real-time.

A Novel CB/PDMS Electrode For ECG Monitoring During Swimming

To collect reliable electrocardiogram (ECG) signals during swimming, waterproof tape must be generously applied to commercial Ag/AgCl electrodes to seal off water penetration. Prolonged use of waterproof tape can lead to skin irritations and when the tape is removed it can tear the skin. PURPOSE: The purpose of the study was to examine if our reusable hydrophobic ECG electrodes, which are composed of carbon black powder/Polydimethylsiloxane (CB/PDMS), can be effectively used to collect reliable ECG signals underwater. METHODS: An ECG signal (Lead I configuration) was collected using a custom-made chest strap which incorporated our hydrophobic electrodes. Two CB/PDMS electrodes were attached to the strap, which also had a miniaturized ECG data acquisition board. The ECG signal and tri-axis accelerometric data were saved directly into a micro-SD card in the ECG circuit board. Three participants were recruited to examine the electrodes’ performance during two types of swimming (freestyle and breaststroke). The experimental protocol included having the subjects stand while not immersed (dry condition, 1 min) and then swim each type of stroke (1 lap each (45 meter)). We examined the ECG signal quality by observing if there were reductions of QRS peak amplitudes during swimming when compared to the dry condition. Specifically, we calculated cross-correlation indices of ECG templates to quantitatively determine possible morphological changes in ECG waveforms when electrodes were exposed to water. Moreover, we computed the amount of signal dropout due to either motion artifacts or water penetration into our electrodes. RESULTS: No significant reduction of QRS peak amplitude was observed between dry and swimming conditions, for both freestyle and breaststroke. The cross correlation indices of ECG templates between dry and swimming conditions for freestyle and breaststroke were both 99.9%. Moreover, there was minimal ECG signal loss (less than 1%) due to either motion artifacts or water penetration. CONCLUSIONS: Our CB/PDMS electrodes can provide high-fidelity ECG signal morphologies without any amplitude degradation and minimal signal dropout during swimming. These electrodes have the potential to be applicable to better understanding underwater physiology, and as a performance-enhancement training aid.