ECG Waveform Research Articles

A Hybrid Energy Efficient Nano Architecture using Memristors Chaos Systems for Secure Communication

With an advent of IoT (Internet of Things), number of networked devices increases day by day and will reach its peak by the end of 2020. IoT plays an important role in smart automation, medical informatics, computer vision and even in wearable device engineering. Since IoT transfers the huge data from the machine to machines, achieving the compact size, low power and overcoming the security breaches remains the real challenge among the researchers. With the integration of Nanotechnology, IoT has reached its new dimension to meet the above constraints but still requires much more lime light of research. To meet the above criteria, the paper proposes the new algorithm HEENA (Hybrid Energy Efficient Nano Architecture) which uses the memristors based modelling for data encryption process. The proposed system overcomes the existing CMOS /Nano transistor-based modelling and has following contributions 1) Design of memristor based chaotic non -linear system 2) Generation of Initial conditions using ECG waveforms 3) Generation of High complex Cipher Keys using Diffusion process. The test bed has been developed using BCM2578 SoC and the proposed architecture has been evaluated on the above test bed in terms of sensitivity, and entropy Moreover, the energy consumption is calculated at various cases and results shows that the proposed architecture consumes the less power along with the secured data communication.

Abstract 456: Fundamental New Clinical Insight From Excitation-contraction Coupling Studies of the Cardiovascular System via Integrated ECG And Photoplethysmography (PPG) Analyses During Sleep

Introduction: The surface ECG provides critical insights into cardiac excitation. The ECG waveform dynamics inform clinical decisions beyond heart rate and QT measures. Likewise, the PPG waveforms contain physiologically rich information about a person’s health, but it’s clinical use is limited to peripheral oxygen saturation. Importantly, the complexity of ECG and PPG dynamics reflect multi-directional interconnected networks of cellular signaling pathways, tissues, organs and organ system interactions that regulate homeostatic function under changing conditions. While integrated dynamic analysis of ECG and PPG is rooted in fundamental principles of electromechanical coupling and systems biology, this has been unexplored for identifying subclinical illness. Hypothesis: Integrated dynamic analysis of ECG and PPG waveforms during sleep identify subclinical cardiovascular illness in a multicenter study of asymptomatic community subjects. Methods & Results: We studied polysomnography recordings from 137 adults in a substudy of the multi-ethnic study of atherosclerosis (MESA-5). The cohort had mean age of 69±10 years, 43% women, 11% African Americans, 48% diabetes mellitus (DM) of which 33% received insulin therapy (DM-RxI), 21% prior myocardial infarction (MI), and 16% ostensibly healthy (i.e., unremarkable findings on history, physical exam, lab, etc). Dynamic analysis of PPG waveforms were performed for each sleep stage: (1) Awake before onset of sleep; (2) Non-REM sleep; (3) REM sleep; (4) Arousals during sleep; (5) Awake after completing sleep. In a multivariate model, the odds ratio (5th compared to 1st quintile) for MI for sleep stage 1 was 7.5 (2.1-27), for DM sleep stage 5 was 0.057 (0.0040-0.82), and for DM-RxI for sleep stage 5 was 0.23 (0.057-0.91), with corresponding ROC curve areas ranging between 0.65-0.90 (p < 0.05). Conclusions: Integrated dynamic analysis of ECG and PPG waveforms provide fundamental new insight for identifying an individual’s health status, including MI, DM and treatment for DM. This novel strategy has potential for broad clinical application and warrants prospective validation.

A Comparison of ECG Signal Pre-processing Using FrFT, FrWT and IPCA for Improved Analysis

ObjectiveElectrocardiogram (ECG) is a diagnostic tool for recording electrical activities of the human heart non-invasively. It is detected by electrodes placed on the surface of the skin in a conductive medium. In medical applications, ECG is used by cardiologists to observe heart anomalies (cardiovascular diseases) such as abnormal heart rhythms, heart attacks, effects of drug dosage on subject's heart and knowledge of previous heart attacks. Recorded ECG signal is generally corrupted by various types of noise/distortion such as cardiac (isoelectric interval, prolonged depolarization and atrial flutter) or extra cardiac (respiration, changes in electrode position, muscle contraction and power line noise). These factors hide the useful information and alter the signal characteristic due to low Signal-to-Noise Ratio (SNR). In such situations, any failure to judge the ECG signal correctly may result in a delay in the treatment and harm a subject (patient) health. Therefore, appropriate pre-processing technique is necessary to improve SNR to facilitate better treatment to the subject. Effects of different pre-processing techniques on ECG signal analysis (based on R-peaks detection) are compared using various Figures of Merit (FoM) such as sensitivity (Se), accuracy (Acc) and detection error rate (DER) along with SNR. MethodsIn this research article, a new fractional wavelet transform (FrWT) has been proposed as a pre-processing technique in order to overcome the disadvantages of other existing commonly used techniques viz. wavelet transform (WT) and the fractional Fourier transform (FrFT). The proposed FrWT technique possesses the properties of multiresolution analysis and represents signal in the fractional domain which consists of representation in terms of rotation of signals in the time–frequency plane. In the literature, ECG signal analysis has been improvised using statistical pre-processing techniques such as principal component analysis (PCA), and independent component analysis (ICA). However, both PCA and ICA are prone to suffer from slight alterations in either signal or noise, unless the basis functions are prepared with a worldwide set of ECG. Independent Principal Component Analysis (IPCA) has been used to overcome this shortcoming of PCA and ICA. Therefore, in this paper three techniques viz. FrFT, FrWT and IPCA are selected for comparison in pre-processing of ECG signals. ResultsThe selected methods have been evaluated on the basis of SNR, Se, Acc and DER of the detected ECG beats. FrWT yields the best results among all the methods considered in this paper; 34.37dB output SNR, 99.98% Se, 99.96% Acc, and 0.036% DER. These results indicate the quality of biology-related information retained from the pre-processed ECG signals for identifying different heart abnormalities. ConclusionCorrect analysis of the acquired ECG signal is the main challenge for cardiologist due to involvement of various types of noises (high and low frequency). Twenty two real time ECG records have been evaluated based on various FoM such as SNR, Se, Acc and DER for the proposed FrWT and existing FrFT and IPCA preprocessing techniques. Acquired real-time ECG database in normal and disease situations is used for the purpose. The values of FoMs indicate high SNR and better detection of R-peaks in a ECG signal which is important for the diagnosis of cardiovascular disease. The proposed FrWT outperforms all other techniques and holds both analytical attributes of the actual ECG signal and alterations in the amplitudes of various ECG waveforms adequately. It also provides signal portrayals in the time-fractional-frequency plane with low computational complexity enabling their use practically for versatile applications.

Miniature Wireless ECG Sensor for Mobile Medical Treatment

A kind of miniature wireless ECG sensor is designed for mobile medicine in the paper.The miniature wireless ECG sensor acquires ECG signals of human body by single lead,and it transmits the processed ECG data to mobile phone through Bluetooth transmission technology to realize ECG waveform,heart rate display and ECG data storage.The wireless ECG sensor has the advantages of simple operation,high accuracy,strong anti-interference ability and good real-time performance.It can monitor some arrhythmia and other diseases for a long time.

765: Intrapartum fetal electrocardiogram (ECG) in small and large-for-gestational age fetuses

Fetal ST-depression and ST-elevation reflect abnormal repolarization and are associated with cardiac hypoxia. Both small-for-gestational age (SGA) and large-for-gestational age (LGA) infants are at increased risk of cardiac dysfunction compared to appropriate-for-gestational age (AGA) infants. The objective of this study was to evaluate whether intrapartum ST-changes occur more frequently in SGA or LGA as compared to AGA infants. We conducted a secondary analysis of a large, multi-centered randomized-controlled trial that took place at 26 academically-affiliated US hospitals. Laboring women with singleton pregnancy at 36 weeks of gestation or greater underwent fetal ECG waveform analysis using the STAN S31 device (Neoventa Medical) and were randomly assigned to masked or unmasked readings. In our analysis, we included subjects with available ECG-tracings and excluded those with diabetes mellitus and major fetal anomalies. A gender and race/ethnicity-specific nomogram was utilized to characterize infant birth weight as SGA (<10%ile), LGA (>90%ile) or AGA (10-90%ile). Our primary outcome was the frequency of fetal ECG tracings with intrapartum ST-changes (ST-depression or ST-elevation without ST-depression), analyzed separately for the first and second stages of labor, among SGA, LGA and AGA infants. Univariable analyses were performed to compare maternal characteristics across groups. All significant covariates (P<0.05) and duration of ECG monitoring were included in multivariable logistic regression models. Our study included 4,971 eligible women in the first stage of labor and 4,074 in the second. Maternal characteristics are displayed in Table 1. In the adjusted model, neither ST-depression nor elevation occurred more frequently in fetal ECG tracings of SGA or LGA compared to AGA births in the first or second stage of labor (Table 2). Intrapartum fetal ECG ST-elevation and depression were not associated with the extremes of fetal growth.View Large Image Figure ViewerDownload Hi-res image Download (PPT)

AF detection from ECG recordings using feature selection, sparse coding, and ensemble learning

Objective: The objective of this paper is to provide an algorithm for accurate, automated detection of atrial fibrillation (AF) from ECG signals. Four types of ECG signals are considered: normal signals, signals representing symptoms of AF, other signals, and noisy signals. This paper represents follow-up work to the authors’ entry in the 2017 PhysioNet Challenge as reported in the 2017 Computing in Cardiology Conference. Approach: Our approach involves extracting features from the ECG waveform and training a machine learning classifier. In feature extraction, we calculate several statistical features related to the ECG signal and fiduciary points. We also used a disciplined method of feature selection to reduce the dimensionality of the feature space. We also employ sparse coding as an unsupervised feature extraction tool. The classifier we use is a decision tree-based ensemble learning classifier. Main results: When applied to the hidden test data reserved by the PhysioNet Challenge organizers, our classifier reports F1 scores of 0.91, 0.78, and 0.71 for the Normal, AF, and Other classes, respectively. The overall test score is 0.80, and is obtained by averaging the F1 scores for these three classes. Significance: This work demonstrates that feature selection and ensemble learning can be used to improve the performance of ECG-based classification of AF.

Abstract 145: Efficacy of ECG Waveform to Predict Outcome in Patients With Out-of-Hospital Cardiac Arrest

Introduction: Neurological intact survival of pulseless electrical activity (PEA) has extremely poor outcome when compared with ventricular fibrillation. In Japan, the all patients with out-of-hospital cardiac arrest received CPR is transferred to emergency hospitals. To determine termination of resuscitation, it is necessary to predict their outcome. Therefore, we analyzed the relationship between ECG changes during prehospital resuscitation and outcome. Method: This study is a retrospective observational study for out-of-hospital PEA in April 2012 - April 2017 which was transported to the emergency medical center in Shizuoka General Hospital. We examined consecutive PEA cases where recorded by emergency medical services. The QRS width and RR interval were measured in ECG recording in the initial check pulse phase and arrival hospital. The primary outcome was survival admission and the secondary outcome was good neurologic outcome defined as cerebral performance score (CPC) of 1, 2. ECG changes determined as QRS width and RR interval in initial rhythm minus those at hospital arrival. Results: Eighty-four patients were analyzed and the average age was 77.3 years old, male 52.0%. There were 44 cases (52%) with survival admission (ROSC cases) and 7 (8%) with CPC 1-2. ROSC cases showed significantly narrower QRS width than non-ROSC cases in the initial waveform (139.1±61.9 vs 165.5±57.5msec, P&lt;0.05). Regarding ECG waveform changes, RR interval was significantly shortened in ROSC cases(-356.0±1256.0 vs 386.8±1258.6msec, P&lt;0.01). Comparing CPC1-2 cases and CPC3-5 cases, QRS width is significantly narrower in CPC1-2 cases(108.6±44.51 vs 155.6±60msec, P&lt;0.05), and there were no significant differences in ECG waveform changes. Multivariate analysis revealed that Initial QRS width and RR interval change are the independent factors associated with survival admission. Conclusion: Those findings suggested that analyzing QRS width and change of RR intervals during prehospital resuscitation predict survival admission in PEA cases.

Aging Alters the Relative Contributions of the Sympathetic and Parasympathetic Nervous System to Blood Pressure Control in Women.

Autonomic support of blood pressure increases with age in humans. Large differences exist in the dose of trimethaphan (TMP) required for ganglionic blockade in young and older women. We asked whether differences in the dose of TMP required to achieve ganglionic blockade are because of differences in the relative contributions of the sympathetic and parasympathetic nervous system in control of blood pressure with age. Muscle sympathetic nerve activity (microneurography, peroneal nerve), heart rate (HR), and blood pressure were recorded before and during incremental doses of TMP camsylate until ganglionic blockade was achieved (absence of muscle sympathetic nerve activity and <5-bpm increase in HR during a valsalva maneuver; final TMP dose, 1-7 mg/min). HR variability was analyzed from the ECG waveform (WinCPRS). The dose of TMP required to achieve ganglionic blockade is positively related to basal HR variability, where women with high HR variability require a higher dose of TMP to achieve ganglionic blockade. In contrast, baseline muscle sympathetic nerve activity is inversely related with the dose of TMP required to achieve ganglionic blockade, such that women with high basal muscle sympathetic nerve activity required a lower dose of TMP. As such, the change in HR with ganglionic blockade was positively related, and the change in mean arterial pressure was inversely related, with the dose of TMP required to achieve ganglionic blockade. These data suggest loss of parasympathetic tone and increased sympathetic tone with aging contribute to the increase in blood pressure with age in women and dictate the dose of TMP that is necessary to achieve ganglionic blockade.

Cardiac Conduction Model for Generating 12 Lead ECG Signals With Realistic Heart Rate Dynamics

We present an extended heterogeneous oscillator model of cardiac conduction system for generation of realistic 12 lead ECG waveforms. The model consists of main natural pacemakers represented by modified van der Pol equations, and atrial and ventricular muscles, in which the depolarization and repolarization processes are described by modified FitzHugh-Nagumo equations. We incorporate an artificial RR-tachogram with the specific statistics of a heart rate, the frequency-domain characteristics of heart rate variability produced by Mayer and respiratory sinus arrhythmia waves, normally distributed additive noise and a baseline wander that couple the respiratory frequency. The standard 12 lead ECG is calculated by means of a weighted linear combination of atria and ventricle signals and thus can be fitted to clinical ECG of real subject. The model is capable to simulate accurately realistic ECG characteristics including local pathological phenomena accounting for biophysical properties of the human heart. All these features provide significant advantages over existing nonlinear cardiac models. The proposed model constitutes a useful tool for medical education and for assessment and testing of ECG signal processing software and hardware systems.

Biomedical Arrhythmia Heart Diseases Classification Based on Artificial Neural Network and Machine Learning Approach

In present day, several types of developments are carried toward the medical application. There has been increased improvement in the processing of ECG signals. The accurate detection of ECG signals with the help of detection of P, Q, R and S waveform. However these waveforms are suffered from some disturbances like noise. Initially denoising the ECG signal using filters and detect the PQRS waveforms. Four filters are carried out to remove the ECG noises that are Median, Gaussian, FIR and Butterworth filter. ECG signal is analyzed or classify using Extreme Learning Machine (ELM) and it compared with Support Vector Machine (SVM) and Back Propagation Neural Network (BPNN). The paper classifies the ECG signal into two classes, Normal and Abnormal. ECG waveform is detected and analyzed using the 48 records of the MIT-BIH arrhythmia database. Denoising results are evaluated using MSE, RMSE, PSNR, NAE and NCC. The classifier performance is measured in terms of Sensitivity (Se), Positive Predictivity (PP) and Specificity (SP).

Hierarchical Probabilistic Framework for Fetal R-Peak Detection, Using ECG Waveform and Heart Rate Information

The abdominal fetal electrocardiogram (fECG) can provide valuable information about fetal well-being. However, fetal R-peak detection in abdominal fECG recordings is challenging due to the low signal-to-noise ratio (SNR) and the nonstationary nature of the fECG waveform in the abdominal recordings. In this paper, we propose a multichannel hierarchical probabilistic framework for fetal R-peak detection that combines predictive models of the ECG waveform and the heart rate. The performance of our method was evaluated on set-A of the 2013 Physionet/Computing in Cardiology Challenge and compared to the performance of several methods that have been proposed in the literature. The hierarchical probabilistic framework presented in this study outperforms other methods for fetal R-peak detection with a mean overall detection accuracy for set-A of 99.6%. Even for recordings with low SNR our method enables reliable fetal R-peak detection (Ac 99.4%).

Photoplethysmographic assessment of pulse transit time correlates with echocardiographic measurement of stroke volume in preterm infants with patent ductus arteriosus.

We aimed to correlate photoplethysmographic parameters with stroke volume in infants with PDA. Photoplethysmography constitutes the optical signal in pulse oximetry. Stroke volume was determined echocardiographically. Pulse transit time, right hand to foot arrival time difference, and relative amplitude were measured from pulse oximeter and ECG waveforms. Photoplethysmographic parameters before and after PDA closure were compared with stroke volume. After PDA closure, pulse transit time to the hand and to the foot were prolonged (54.7 ± 6.7 vs 65.5 ± 9.8 ms, p < 0.001, 82.5 ± 12.8 vs 88.6 ± 10.6 ms, p = 0.03), arrival time difference decreased (27.7 ± 7.6 vs 23.1 ± 5.6 ms, p = 0.021), and relative amplitude decreased (from 2.1 ± 0.7% to 1.5 ± 0.5%, p = 0.003). The time-based photoplethysmographic parameters correlated with stroke volume. Photoplethysmographic waveform parameters are significantly different before and after PDA closure and the time-based parameters correlate well with stroke volume. Monitoring pulse transit time may assist in evaluation for spontaneous PDA closure or response to therapy.

A Fiducial Scaffold for ECG Compression in Low-Powered Devices.

Large volumes of physiological data can now be routinely collected using wearable devices, though a key challenge that remains is the conversion of raw data into clinically relevant and actionable information. While power constraints prevent continuous wireless streaming of large amounts of raw data for offline processing, on-board microprocessors have become sufficiently powerful for data reduction to be performed in real time on the wearable device itself, so that only aggregate, clinically interpretable measures need to be transmitted wirelessly. Here, we use the curve-length transform to extract key beat-by-beat information from the raw ECG waveform, and to identify clinically relevant timing and amplitude information. Each beat is parameterized by 12 morphological features that serve as fiducial markers, sufficient to directly reconstruct a scaffold representation of the ECG waveform. At a nominal heart rate of 70 beats/min and a sampling rate of 250 Hz, typical for wearable monitors, this represents approximately an 18-fold compression. Using difference encoding, the compression ratio improves to 21. Our algorithm computes a running exponentially-weighted average of each identified morphological feature. When any feature deviates significantly from its running average, the algorithm retains the raw waveform for five beats preceding and following the anomaly, enabling future review of the raw data. The algorithm automatically located 93.8% of the 3,615 expert-annotated QRS onsets and offsets in the PhysioNet QT-Database to within 20 ms. Similarly, it located 83.5% of all 3,194 P-wave onset and offset annotations to within 32 ms, and 89.0% of all 3,542 T-wave offset annotations within 72 ms.

Real Time Detection of R – Peak in QRS Complex of ECG using Microcontroller

&lt;p&gt;Real-time detection of R peaks in QRS complex of ECG signal is the first step in the processing of ECG waveform. Based on this, various other ECG parameters can be extracted. These parameters provide substantial information about various heart diseases. In this paper, we are proposing a method to detect R – peaks of ECG signal dynamically. The most prominent role in the R – peak detector is executed by the microcontroller. This method originates by acquiring signal from the subject and necessary pre-processing is carried out on the signal in order to achieve the denoised signal. Subsequently, this filtered signal is handed over to microcontroller where a pulse is generated for each R – peak that is found in the QRS complex of ECG signal. The microcontroller is embedded with a signal processing algorithm. The algorithm used to determine the R – peaks is double differentiation method which is straightforward and robust. &lt;/p&gt;

Arrhythmia Recognition and Classification Using ECG Morphology and Segment Feature Analysis.

In this work, arrhythmia appearing with the presence of abnormal heart electrical activity is efficiently recognized and classified. A novel method is proposed for accurate recognition and classification of cardiac arrhythmias. Firstly, P-QRS-T waves is segmented from ECG waveform; secondly, morphological features are extracted from P-QRS-T waves, and ECG segment features are extracted from the selected ECG segment by using PCA and dynamic time warping(DTW); finally, SVM is applied to the features and automatic diagnosis results is presented. ECG data set used is derived from the MIT-BIH in which ECG signals are divided into the four classes: normal beats(N), supraventricular ectopic beats (SVEBs), ventricular ectopic beats (VEBs) and fusion of ventricular and normal (F). Our proposed method can distinguish N, SVEBs, VEBs and F with an accuracy of 97.80 percent. The sensitivities for the classes N, SVEBs, VEBs and F are 99.27, 87.47, 94.71, and 73.88 percent and the positive predictivities are 98.48, 95.25, 95.22 and 86.09 percent respectively. The detection sensitivity of SVEBs and VEBs has a better performance by combining proposed features than by using the ECG morphology or ECG segment features separately. The proposed method is compared with four selected peer algorithms and delivers solid results.

Czy jest możliwa automatyczna interpretacja EKG za pomocą zdjęcia wykonanego smartfonem?

Introduction. Nowadays there are no tools enabling fast digitalization and supporting interpretation of paper ECG recordings. Popularity and availability of devices like smartphone could be used in clinical practice as tools enhancing the interpretation of ECG recordings. The aim of the study is to determine possibilities of using free mobile application and smartphone in assessment of QRS frequency and electrical axis of the heart as a basis for more advanced analysis. Materials and methods. Fifty recordings of 12-lead ECG at 25 mm/s generated by devices of various producers were qualified for the analysis. Each of the recordings was assessed as diagnostic by two cardiologists, who also measured the frequency of QRS complexes, electrical axis of the heart, duration of QT, amplitude and duration of QRS complexes and T waves. Afterwards, automatic interpretation of ECG recordings was performed with educational mobile application eEKG (available for free at AppStore and Google Play) and iPhone 5s. The pictures of ECG waveforms were taken according to instruction of the application. The results of expert assessment and automatic interpretation were compared. The 10% acceptable margin of error was established for assessment of frequency of QRS complexes by the application. Results. Six hundred ECG waveforms (12 leads in every ECG recording) were analysed for frequency of QRS complexes and 50 ECG recordings were analyzed for electrical axis of the heart. The application qualified as diagnostic 573 (95.5%) attempts of QRS frequency assessment and 26 (52%) attempts of electrical axis of the heart assessment. The assessment was accurate in 82% of attempts for QRS complexes frequency assessment and in 96% of attempts for electrical axis of the heart assessment. Significant correlation was proven between QT duration, T wave amplitude, ratio of amplitudes of QRS complexes and T waves and effectiveness of automatic interpretation of ECG waveform. Conclusions. Effective digitalisation and automatic interpretation of ECG recording in assessment of frequency of QRS complexes and electrical axis of the heart is possible with mobile application and smartphone type device.

Effects of intra-resuscitation antiarrhythmic administration on rearrest occurrence and intra-resuscitation ECG characteristics in the ROC ALPS trial

BackgroundIntra-resuscitation antiarrhythmic drugs may improve resuscitation outcomes, in part by avoiding rearrest, a condition associated with poor out-of-hospital cardiac arrest (OHCA) outcomes. However, antiarrhythmics may also alter defibrillation threshold. The objective of this study was to investigate the relationship between rearrest and intra-resuscitation antiarrhythmic drugs in the context of the Resuscitation Outcomes Consortium (ROC) amiodarone, lidocaine, and placebo (ALPS) trial. HypothesisRearrest rates would be lower in cases treated with amiodarone or lidocaine, versus saline placebo, prior to first return of spontaneous circulation (ROSC). We also hypothesized antiarrhythmic effects would be quantifiable through analysis of the prehospital electrocardiogram. MethodsWe conducted a secondary analysis of the ROC ALPS trial. Cases that first achieved prehospital ROSC after randomized administration of study drug were included in the analysis. Rearrest, defined as loss of pulses following ROSC, was ascertained from emergency medical services records. Rearrest rate was calculated overall, as well as by ALPS treatment group. Multivariable logistic regression models were constructed to assess the association between treatment group and rearrest, as well as rearrest and both survival to hospital discharge and survival with neurologic function. Amplitude spectrum area, median slope, and centroid frequency of the ventricular fibrillation (VF) ECG were calculated and compared across treatment groups. ResultsA total of 1144 (40.4%) cases with study drug prior to first ROSC were included. Rearrest rate was 44.0% overall; 42.9% for placebo, 45.7% for lidocaine, and 43.0% for amiodarone. In multivariable logistic regression models, ALPS treatment group was not associated with rearrest, though rearrest was associated with poor survival and neurologic outcomes. AMSA and median slope measures of the first available VF were associated with rearrest case status, while median slope and centroid frequency were associated with ALPS treatment group. ConclusionRearrest rates did not differ between antiarrhythmic and placebo treatment groups. ECG waveform characteristics were correlated with treatment group and rearrest. Rearrest was inversely associated with survival and neurologic outcomes.

Atrial fibrillation classification using step-by-step machine learning

This paper presents a detailed overview of our submission to the 2017 Physionet Challenge where competitors were asked to build a model to classify a single lead ECG waveform as either normal sinus rhythm, atrial fibrillation, other rhythm, or noisy. A step-by-step machine learning pipeline was assembled, which included signal conditioning, R-peak detection and filtering, and feature extraction. A suite of over 300 features, falling into one of three main feature groups; template features, RRI features, and full waveform features, were extracted from each waveform and an XGBoost, tree-based, gradient boosting classifier was used as the machine learning algorithm. The model produced a cross-validation F1 score of 0.8245, a hidden sub-test score of 0.82, and a hidden test score of 0.8125. The score breakdown for each class (normal sinus rhythm, atrial fibrillation, other rhythm, and noisy) was as follows: F1, NRS = 0.9024, F1, AF = 0.8156, F1, OR = 0.7194, F1, Noise = 0.5705.

Variability in the time to initiation of CPR in continuously monitored pediatric ICUs

AimTo study the influence of patient characteristics and unit ergonomics and human factors on the time to initiation of CPR. MethodsA single center study of children, 0 to 21 years old, admitted to an ICU who experienced cardiopulmonary arrest (CPA) requiring >1 min of chest compressions. Time of CPA was determined by analysis of continuous ECG, plethysmography, arterial blood pressure, and end-tidal CO2 (EtCO2) waveforms. Initiation of CPR was identified by the onset of cyclic artifact in the ECG waveform. Patient characteristics and unit ergonomics and human factors were examined including CPA cause, identification on the High-Risk Checklist (HRC), existing monitoring, ICU type, time of day, nursing shift change, and outcome. ResultsThe median time from CPA to initiation of CPR was 50.5 s (IQR 26.5 to 127.5) in 36 CPAs. Forty-seven percent of patients experienced time from CPA to initiation of CPR of >1 min. There was no difference in CPA cause, ICU type, time of day, or nursing shift change. ConclusionNearly half of pediatric patients who experienced CPA in an ICU setting did not meet AHA guidelines for early initiation of CPR. This is an opportunity to study the recognition phase of CPA using continuous monitoring data with the aim of improving the understanding of and factors contributing to delays in initiation of CPR.

Study on a Biometric Authentication Model based on ECG using a Fuzzy Neural Network

Traditional authentication methods use numbers or graphic passwords and thus involve the risk of loss or theft. Various studies are underway regarding biometric authentication because it uses the unique biometric data of a human being. Biometric authentication technology using ECG from biometric data involves signals that record electrical stimuli from the heart. It is difficult to manipulate and is advantageous in that it enables unrestrained measurements from sensors that are attached to the skin. This study is on biometric authentication methods using the neural network with weighted fuzzy membership functions (NEWFM). In the biometric authentication process, normalization and the ensemble average is applied during preprocessing, characteristics are extracted using Haar-wavelets, and a registration process called “training” is performed in the fuzzy neural network. In the experiment, biometric authentication was performed on 73 subjects in the Physionet Database. 10-40 ECG waveforms were tested for use in the registration process, and 15 ECG waveforms were deemed the appropriate number for registering ECG waveforms. 1 ECG waveforms were used during the authentication stage to conduct the biometric authentication test. Upon testing the proposed biometric authentication method based on 73 subjects from the Physionet Database, the TAR was 98.32% and FAR was 5.84%.