Enhancing the accuracy of shock advisory algorithms in automated external defibrillators during ongoing cardiopulmonary resuscitation using a cascade of CNNEDs.

Delivery of continuous cardiopulmonary resuscitation (CPR) plays an important role in the out-of-hospital cardiac arrest (OHCA) survival rate. However, to prevent CPR artifacts being superimposed on ECG morphology data, currently available automated external defibrillators (AEDs) require pauses in CPR for accurate analysis heart rhythms. In this study, we propose a novel Convolutional Neural Network-based Encoder-Decoder (CNNED) structure with a shock advisory algorithm to improve the accuracy and reliability of shock versus non-shock decision-making without CPR pause in OHCA scenarios. Our approach employs a cascade of CNNEDs in conjunction with an AED shock advisory algorithm to process the ECG data for shock decisions. Initially, a CNNED trained on an equal number of shockable and non-shockable rhythms is used to filter the CPR-contaminated data. The resulting filtered signal is then fed into a second CNNED, which is trained on imbalanced data more tilted toward the specific rhythm being analyzed. A reliable shock versus non-shock decision is made when both classifiers from the cascade structure agree, while segments with conflicting classifications are labeled as indeterminate, indicating the need for additional segments to analyze. To evaluate our approach, we generated CPR-contaminated ECG data by combining clean ECG data with 52 CPR samples. We used clean ECG data from the CUDB, AFDB, SDDB, and VFDB databases, to which 52 CPR artifact cases were added, while a separate test set provided by the AED manufacturer Defibtech LLC was used for performance evaluation. The test set comprised 20,384 non-shockable CPR-contaminated segments from 392 subjects, as well as 3744 shockable CPR-contaminated samples from 41 subjects with coarse ventricular fibrillation (VF) and 31 subjects with rapid ventricular tachycardia (rapid VT). We observed improvements in rhythm analysis using our proposed cascading CNNED structure when compared to using a single CNNED structure. Specifically, the specificity of the proposed cascade of CNNED structure increased from 99.14% to 99.35% for normal sinus rhythm and from 96.45% to 97.22% for other non-shockable rhythms. Moreover, the sensitivity for shockable rhythm detection increased from 90.90% to 95.41% for ventricular fibrillation and from 82.26% to 87.66% for rapid ventricular tachycardia. These results meet the performance thresholds set by the American Heart Association and demonstrate the reliable and accurate analysis of heart rhythms during CPR using only ECG data without the need for CPR interruptions or a reference signal.

Deep Neural Network Approach for Continuous ECG-Based Automated External Defibrillator Shock Advisory System During Cardiopulmonary Resuscitation.

Background Because chest compressions induce artifacts in the ECG, current automated external defibrillators instruct the user to stop cardiopulmonary resuscitation (CPR) while an automated rhythm analysis is performed. It has been shown that minimizing interruptions in CPR increases the chance of survival. Methods and Results The objective of this study was to apply a deep-learning algorithm using convolutional layers, residual networks, and bidirectional long short-term memory method to classify shockable versus nonshockable rhythms in the presence and absence of CPR artifact. Forty subjects' data from Physionet with 1131 shockable and 2741 nonshockable samples contaminated with 43 different CPR artifacts that were acquired from a commercial automated external defibrillator during asystole were used. We had separate data as train and test sets. Using our deep neural network model, the sensitivity and specificity of the shock versus no-shock decision for the entire data set over the 4-fold cross-validation sets were 95.21% and 86.03%, respectively. This result was based on the training and testing of the model using ECG data in both the presence and the absence of CPR artifact. For ECG without CPR artifact, the sensitivity was 99.04% and the specificity was 95.2%. A sensitivity of 94.21% and a specificity of 86.14% were obtained for ECG with CPR artifact. In addition to 4-fold cross-validation sets, we also examined leave-one-subject-out validation. The sensitivity and specificity for the case of leave-one-subject-out validation were 92.71% and 97.6%, respectively. Conclusions The proposed trained model can make shock versus nonshock decision in automated external defibrillators, regardless of CPR status. The results meet the American Heart Association's sensitivity requirement (>90%).

Detection of low-frequency oscillations in renal blood flow.

Detection of the low-frequency (LF; approximately 0.01 Hz) component of renal blood flow, which is theorized to reflect the action of a third renal autoregulatory mechanism, has been difficult due to its slow dynamics. In this work, we used three different experimental approaches to detect the presence of the LF component of renal autoregulation using normotensive and spontaneously hypertensive rats (SHR), both anesthetized and unanesthetized. The first experimental approach utilized a blood pressure forcing in the form of a chirp, an oscillating perturbation with linearly increasing frequency, to elicit responses from the LF autoregulatory component in anesthetized normotensive rats. The second experimental approach involved collection and analysis of spontaneous blood flow fluctuation data from anesthetized normotensive rats and SHR to search for evidence of the LF component in the form of either amplitude or frequency modulation of the myogenic and tubuloglomerular feedback mechanisms. The third experiment used telemetric recordings of arterial pressure and renal blood flow from normotensive rats and SHR for the same purpose. Our transfer function analysis of chirp signal data yielded a resonant peak centered at 0.01 Hz that is greater than 0 dB, with the transfer function gain attenuated to lower than 0 dB at lower frequencies, which is a hallmark of autoregulation. Analysis of the data from the second experiments detected the presence of approximately 0.01-Hz oscillations only with isoflurane, albeit at a weaker strength compared with telemetric recordings. With the third experimental approach, the strength of the LF component was significantly weaker in the SHR than in the normotensive rats. In summary, our detection via the amplitude modulation approach of interactions between the LF component and both tubuloglomerular feedback and the myogenic mechanism, with the LF component having an identical frequency to that of the resonant gain peak, provides evidence that 0.01-Hz oscillations may represent the third autoregulatory mechanism.

OxiMA: A Frequency-Domain Approach to Address Motion Artifacts in Photoplethysmograms for Improved Estimation of Arterial Oxygen Saturation and Pulse Rate.

OBJECTIVE: The purpose of this paper is to demonstrate that a new algorithm for estimating arterial oxygen saturation can be effective even with data corrupted by motion artifacts (MAs). METHODS: OxiMA, an algorithm based on the time-frequency components of a photoplethysmogram (PPG), was evaluated using 22-min datasets recorded from 10 subjects during voluntarily-induced hypoxia, with and without subject-induced MAs. A Nellcor OxiMax transmission sensor was used to collect an analog PPG while reference oxygen saturation and pulse rate (PR) were collected simultaneously from an FDA-approved Masimo SET Radical RDS-1 pulse oximeter. RESULTS: The performance of our approach was determined by computing the mean relative error between the PR/oxygen saturation estimated by OxiMA and the reference Masimo oximeter. The average estimation error using OxiMA was 3 beats/min for PR and 3.24% for oxygen saturation, respectively. CONCLUSION: The results show that OxiMA has great potential for improving the accuracy of PR and oxygen saturation estimation during MAs. SIGNIFICANCE: This is the first study to demonstrate the feasibility of a reconstruction algorithm to improve oxygen saturation estimates on a dataset with MAs and concomitant hypoxia.

Time-varying causal coherence function and its application to renal blood pressure and blood flow data.

This paper describes the development of a model-based approach to estimating both feedforward and feedback paths of causal time-varying coherence functions (TVCF). Theoretical derivations of the coherence bounds of the causal TVCF using the proposed approach are also provided. Both theoretical derivations and simulation results revealed interesting observations, and they were corroborated using experimental renal blood pressure and flow data. Specifically, both theoretical derivations and experimental data showed that in certain cases, the calculation of the traditional TVCF was inappropriate when the system under investigation was a causal system. Moreover, the use of the causal TVCF not only provides quantitative assessment of the coupling between the two signals, but it also provides valuable insights into the composition of the physical structure of the renal autoregulatory system.

An approach to estimate time-varying casual coherence function.

OBJECTIVES: This paper describes the development of a model-based approach to estimating both open-loop and causal time-varying coherence functions (TVCF). Theoretical derivations of the coherence bounds using the proposed approach are also provided. METHODS: A time-varying vector autoregressive (VAR) model was used to estimate both open-loop and causal TVCF. The time-varying optimal parameter search method was employed to identify the time-varying model coefficients as well as the model order of the VAR model. RESULTS: Simulation results revealed interesting observations, and they were corroborated using experimental renal blood pressure and flow data. Specifically, experimental data showed that in certain cases, the calculation of the open-loop TVCF might provide incorrect interpretation of the results when the system under investigation was a closed-loop system, which is consistent with theoretical derivations. CONCLUSIONS: The use of the closed-loop TVCF not only provides quantitative assessment of the coupling between the two signals, but it also provides valuable insights into the composition of the physical structure of the system.

Separation of heart rate variability components of the autonomic nervous system by utilizing principal dynamic modes.

This work introduces a modified Principal Dynamic Modes (PDM) methodology using eigenvalue/eigenvector analysis to separate individual components of the sympathetic and parasympathetic nervous contributions to heart rate variability. We have modified the PDM technique to be used with even a single output signal of heart rate variability data, whereas the original PDMs required both input and output data. This method specifically accounts for the inherent nonlinear dynamics of heart rate control, which the current method of power spectrum density (PSD) is unable to do. Propranolol and atropine were administered to normal human volunteers intravenously to inhibit the sympathetic and parasympathetic activities, respectively. With separate applications of the respective drugs, we found a significant decrease in the amplitude of the waveforms that correspond to each nervous activity. Furthermore, we observed near complete elimination of these dynamics when both drugs were given to the subjects. Comparison of our method to the conventional low/high frequency ratio of PSD shows that PDM methodology provides much more accurate assessment of the autonomic nervous balance by separation of individual components of the autonomic nervous activities. The PDM methodology is expected to have an added benefit that diagnosis and prognostication of a patient's health can be determined simply via a non-invasive electrocardiogram.

Accurate identification of periodic oscillations buried in white or colored noise using fast orthogonal search.

We use a previously introduced fast orthogonal search algorithm to detect sinusoidal frequency components buried in either white or colored noise. We show that the method outperforms the correlogram, modified covariance autoregressive (MODCOVAR) and multiple-signal classification (MUSIC) methods. Fast orthogonal search method achieves accurate detection of sinusoids even with signal-to-noise ratios as low as -10 dB, and is superior at detecting sinusoids buried in 1/f noise. Since the utilized method accurately detects sinusoids even under colored noise, it can be used to extract a 1/f noise process observed in physiological signals such as heart rate and renal blood pressure and flow data.

Linear and nonlinear ARMA model parameter estimation using an artificial neural network.

This paper addresses parametric system identification of linear and nonlinear dynamic systems by analysis of the input and output signals. Specifically, we investigate the relationship between estimation of the system using a feedforward neural network model and estimation of the system by use of linear and nonlinear autoregressive moving-average (ARMA) models. By utilizing a neural network model incorporating a polynomial activation function, we show the equivalence of the artificial neural network to the linear and nonlinear ARMA models. We compare the parameterization of the estimated system using the neural network and ARMA approaches by utilizing data generated by means of computer simulations. Specifically, we show that the parameters of a simulated ARMA system can be obtained from the neural network analysis of the simulated data or by conventional least squares ARMA analysis. The feasibility of applying neural networks with polynomial activation functions to the analysis of experimental data is explored by application to measurements of heart rate (HR) and instantaneous lung volume (ILV) fluctuations.

A new algorithm for linear and nonlinear ARMA model parameter estimation using affine geometry.

A linear and nonlinear autoregressive (AR) moving average (MA) (ARMA) identification algorithm is developed for modeling time series data. The new algorithm is based on the concepts of affine geometry in which the salient feature of the algorithm is to remove the linearly dependent ARMA vectors from the pool of candidate ARMA vectors. For noiseless time series data with a priori incorrect model-order selection, computer simulations show that accurate linear and nonlinear ARMA model parameters can be obtained with the new algorithm. Many algorithms, including the fast orthogonal search (FOS) algorithm, are not able to obtain correct parameter estimates in every case, even with noiseless time series data, because their model-order search criteria are suboptimal. For data contaminated with noise, computer simulations show that the new algorithm performs better than the FOS algorithm for MA processes, and similarly to the FOS algorithm for ARMA processes. However, the computational time to obtain the parameter estimates with the new algorithm is faster than with FOS. Application of the new algorithm to experimentally obtained renal blood flow and pressure data show that the new algorithm is reliable in obtaining physiologically understandable transfer function relations between blood pressure and flow signals.

A dual-input nonlinear system analysis of autonomic modulation of heart rate.

Linear analyses of fluctuations in heart rate and other hemodynamic variables have been used to elucidate cardiovascular regulatory mechanisms. The role of nonlinear contributions to fluctuations in hemodynamic variables has not been fully explored. This paper presents a nonlinear system analysis of the effect of fluctuations in instantaneous lung volume (ILV) and arterial blood pressure (ABP) on heart rate (HR) fluctuations. To successfully employ a nonlinear analysis based on the Laguerre expansion technique (LET), we introduce an efficient procedure for broadening the spectral content of the ILV and ABP inputs to the model by adding white noise. Results from computer simulations demonstrate the effectiveness of broadening the spectral band of input signals to obtain consistent and stable kernel estimates with the use of the LET. Without broadening the band of the ILV and ABP inputs, the LET did not provide stable kernel estimates. Moreover, we extend the LET to the case of multiple inputs in order to accommodate the analysis of the combined effect of ILV and ABP effect on heart rate. Analyzes of data based on the second-order Volterra-Wiener model reveal an important contribution of the second-order kernels to the description of the effect of lung volume and arterial blood pressure on heart rate. Furthermore, physiological effects of the autonomic blocking agents propranolol and atropine on changes in the first- and second-order kernels are also discussed.

On the efficacy of linear system analysis of renal autoregulation in rats.

In order to assess the linearity of the mechanisms subserving renal blood flow autoregulation, broad-band arterial pressure fluctuations at three different power levels were induced experimentally and the resulting renal blood flow responses were recorded. Linear system analysis methods were applied in both the time and frequency domain. In the frequency domain, spectral estimates employing FFT, autoregressive moving average (ARMA) and moving average (MA) methods were used; only the MA model showed two vascular control mechanisms active at 0.02-0.05 Hz and 0.1-0.18 Hz consistent with previous experimental findings [Holstein-Rathlou et al., Amer. J. Physiol., vol. 258, 1990.]. In the time domain, impulse response functions obtained from the MA model indicated likewise the presence of these two vascular control mechanisms, but the ARMA model failed to show any vascular control mechanism at 0.02-0.05 Hz. The residuals (i.e., model prediction errors) of the MA model were smaller than the ARMA model for all levels of arterial pressure forcings. The observed low coherence values and the significant model residuals in the 0.02-0.05 Hz frequency range suggest that the tubuloglomerular feedback (TGF) active in this frequency range is a nonlinear vascular control mechanism. In addition, experimental results suggest that the operation of the TGF mechanism is more evident at low/moderate pressure fluctuations and becomes overwhelmed when the arterial pressure forcing is too high.

Nonlinear system analysis of renal autoregulation in normotensive and hypertensive rats.

We compared the dynamic characteristics in renal autoregulation of blood flow of normotensive Sprague-Dawley rats (SDR) and spontaneously hypertensive rats (SHR), using both linear and nonlinear systems analysis. Linear analysis yielded only limited information about the differences in dynamics between SDR and SHR. The predictive ability, as determined by normalized mean-square errors (NMSE), of a third-order Volterra model is better than for a linear model. This decrease in NMSE with a third-order model from that of a linear model is especially evident at frequencies below 0.2 Hz. Furthermore, NMSE are significantly higher in SHR than SDR, suggesting a more complex nonlinear system in SHR. The contribution of the third-order kernel in describing the dynamics of renal autoregulation in arterial blood pressure and blood flow was found to be important. Moreover, we have identified the presence of nonlinear interactions between the oscillatory components of the myogenic mechanism and tubuloglomerular feedback (TGF) at the level of whole kidney blood flow in SDR. An interaction between these two mechanisms had previously been revealed for SDR only at the single nephron level. However, nonlinear interactions between the myogenic and TGF mechanisms are not detected for SHR.

Detection of "noisy" chaos in a time series.

Time series from biological system often displays fluctuations in the measured variables. Much effort has been directed at determining whether this variability reflects deterministic chaos, or whether it is merely "noise". The output from most biological systems is probably the result of both the internal dynamics of the systems, and the input to the system from the surroundings. This implies that the system should be viewed as a mixed system with both stochastic and deterministic components. We present a method that appears to be useful in deciding whether determinism is present in a time series, and if this determinism has chaotic attributes. The method relies on fitting a nonlinear autoregressive model to the time series followed by an estimation of the characteristic exponents of the model over the observed probability distribution of states for the system. The method is tested by computer simulations, and applied to heart rate variability data.

Comparative nonlinear modeling of renal autoregulation in rats: Volterra approach versus artificial neural networks.

Volterra models have been increasingly popular in modeling studies of nonlinear physiological systems. In this paper, feedforward artificial neural networks with two types of activation functions (sigmoidal and polynomial) are utilized for modeling the nonlinear dynamic relation between renal blood pressure and flow data, and their performance is compared to Volterra models obtained by use of the leading kernel estimation method based on Laguerre expansions. The results for the two types of artificial neural networks (sigmoidal and polynomial) and the Volterra models are comparable in terms of normalized mean-square error (NMSE) of the respective output prediction for independent testing data. However, the Volterra models obtained via the Laguerre expansion technique achieve this prediction NMSE with approximately half the number of free parameters relative to either neural-network model. Nonetheless, both approaches are deemed effective in modeling nonlinear dynamic systems and their cooperative use is recommended in general, since they may exhibit different strengths and weaknesses depending on the specific characteristics of each application.

Detection of chaotic determinism in time series from randomly forced maps.

Time series from biological system often display fluctuations in the measured variables. Much effort has been directed at determining whether this variability reflects deterministic chaos, or whether it is merely "noise". Despite this effort, it has been difficult to establish the presence of chaos in time series from biological sytems. The output from a biological system is probably the result of both its internal dynamics, and the input to the system from the surroundings. This implies that the system should be viewed as a mixed system with both stochastic and deterministic components. We present a method that appears to be useful in deciding whether determinism is present in a time series, and if this determinism has chaotic attributes, i.e., a positive characteristic exponent that leads to sensitivity to initial conditions. The method relies on fitting a nonlinear autoregressive model to the time series followed by an estimation of the characteristic exponents of the model over the observed probability distribution of states for the system. The method is tested by computer simulations, and applied to heart rate variability data.

Respiratory rate extraction via an autoregressive model using the optimal parameter search criterion.

We present an autoregressive model-based method which enables accurate respiratory rate extraction from pulse oximeter recordings over a wide range: 12-48 breaths/min. The method uses the optimal parameter search (OPS) technique to estimate accurate AR parameters which are then factorized into multiple pole terms. The pole with the highest magnitude is shown to correspond to the respiratory rate. The performance of the proposed method to extract respiratory rate is compared to the widely used Burg algorithm using both simulation examples and pulse oximeter recordings. In a previous study, we demonstrated several nonparametric time-frequency approaches that were more accurate than Burg's algorithm when the data length was 1 min [Chon, K. H., S. Dash, and K. Ju. IEEE Trans. Biomed. Eng. 56(8):2054-2063, 2009]. One of the key advantages of the AR method is that a shorter data length can be used. Thus, in this study, we reduced the data length to 30 s and applied our OPS algorithm to examine if accurate respiratory rates can be extracted directly from pulse oximeter recordings. It was found that our proposed method's accuracy was consistently better with smaller variance than Burg's method. In particular, our proposed method's accuracy was significantly greater when respiratory rates were lower than 24 breaths/min.

Time-varying methods for characterizing nonstationary dynamics of physiological systems.

BACKGROUND: Accurate and early diagnosis of various diseases and pathological conditions require analysis techniques that can capture time-varying (TV) dynamics. In the pursuit of promising TV signal processing methods applicable to real-time clinical monitoring applications, nonstationary spectral techniques are of great significance. OBJECTIVES: We present two potential practical applications of such techniques in quantifying TV physiological dynamics concealed in photoplethysmography (PPG) signals: early detection of blood-volume loss using a nonparametric approach known as variable frequency complex demodulation (VFCDM), and accurate detection of abrupt changes in respiratory rates using a parametric approach known as combined optimal parameter search and multiple mode particle filtering (COPS-MPF). METHODS: The VFCDM technique has been tested using ear-PPG signals in two study models: mechanically ventilated patients undergoing surgery in operating room settings and spontaneously breathing conscious healthy subjects subjected to lower body negative pressure (LBNP) in laboratory settings. Extraction of respiratory rates has been tested using COPS-MPF technique in finger-PPG signals collected from healthy volunteers with abrupt changes in respiratory rate ranging from 0.1 to 0.4 Hz. RESULTS: VFCDM method showed promise to detect the blood loss noninvasively in mechanical ventilated patients well before blood losses become apparent to the physician. In spontaneously breathing subjects during LBNP experiments, the early detection and quantification of blood loss was possible at 40% of LBNP tolerance. COPS-MPF showed high accuracy in detecting the constant as well as sudden changes in respiratory rates as compared to other time-invariant methods. CONCLUSION: Integration of such robust algorithms into pulse oximeter device may have significant impact in real-time clinical monitoring and point-of-care healthcare settings.

Electrohydraulic pump-driven closed-loop blood pressure-regulatory system.

In this paper, we describe our design for a new electrohydraulic (EH) pump-driven renal perfusion pressure (RPP)-regulatory system capable of implementing precise and rapid RPP regulation in experimental animals. Without this automated system, RPP is manually controlled via a blood pressure clamp, and the imprecision in this method leads to compromised RPP data. This motivated us to develop an EH pump-driven closed-loop blood pressure regulatory system based on flow-mediated occlusion using the vascular occlusive cuff technique. A closed-loop servo-controller system based on a proportional plus integral (PI) controller was designed using the dynamic feedback RPP signal from animals. In vivo performance was evaluated via flow-mediated RPP occlusion, maintenance, and release responses during baseline and ANG II-infused conditions. A step change of -30 mmHg, referenced to normal RPP, was applied to Sprague-Dawley rats with the proposed system to assess the performance of the PI controller. The PI's performance was compared against manual control of blood pressure clamp to regulate RPP. Rapid RPP occlusion (within 3 s) and a release time of approximately 0.3 s were obtained for the PI controller for both baseline and ANG II infusion conditions, in which the former condition was significantly better than manual control. We concluded that the proposed EH RPP-regulatory system could fulfill in vivo needs to study various pressure-flow relationships in diverse fields of physiology, in particular, studying the dynamics of the renal autoregulatory mechanisms.