Personalization of a cardiac electrophysiology model using optical mapping and MRI for prediction of changes with pacing

Computer models of cardiac electrophysiology (EP) can be a very efficient tool to better understand the mechanisms of arrhythmias. Quantitative adjustment of such models to experimental data (personalization) is needed in order to test their realism and predictive power, but it remains challenging at the organ scale. In this paper, we propose a framework for the personalization of a 3-D cardiac EP model, the Mitchell-Schaeffer (MS) model, and evaluate its volumetric predictive power under various pacing scenarios. The personalization was performed on ex vivo large porcine healthy hearts using diffusion tensor MRI (DT-MRI) and optical mapping data. The MS model was simulated on a 3-D mesh incorporating local fiber orientations, built from DT-MRI. The 3-D model parameters were optimized using features such as 2-D epicardial depolarization and repolarization maps, extracted from the optical mapping. We also evaluated the sensitivity of our personalization framework to different pacing locations and showed results on its robustness. Further, we evaluated volumetric model predictions for various epi- and endocardial pacing scenarios. We demonstrated promising results with a mean personalization error around 5 ms and a mean prediction error around 10 ms (5% of the total depolarization time). Finally, we discussed the potential translation of such work to clinical data and pathological hearts.     

The feasibility of endocardial propagation mapping using magnetic resonance guidance in a Swine model, and comparison with standard electroanatomic mapping

The introduction of electroanatomic mapping (EAM) has improved the understanding of the substrate of ventricular tachycardia. EAM systems are used to delineate scar regions responsible for the arrhythmia by creating voltage or activation time maps. Previous studies have identified the benefits of creating MR-guided voltage maps; however, in some cases voltage maps may not identify regions of slow propagation that can cause the reentrant tachycardia. In this study, we obtained local activation time maps and analyzed propagation properties by performing MR-guided mapping of the porcine left ventricle while pacing from the right ventricle. Anatomical and myocardial late gadolinium enhancement images were used for catheter navigation and identification of scar regions. Our MR-guided mapping procedure showed qualitative correspondence to conventional clinical EAM systems in healthy pigs and demonstrated altered propagation in endocardial infarct models.     

Noninvasive computational imaging of cardiac electrophysiology for 3-D infarct

Myocardial infarction (MI) creates electrophysiologically altered substrates that are responsible for ventricular arrhythmias, such as tachycardia and fibrillation. The presence, size, location, and composition of infarct scar bear significant prognostic and therapeutic implications for individual subjects. We have developed a statistical physiological model-constrained framework that uses noninvasive body-surface-potential data and tomographic images to estimate subject-specific transmembrane-potential (TMP) dynamics inside the 3-D myocardium. In this paper, we adapt this framework for the purpose of noninvasive imaging, detection, and quantification of 3-D scar mass for postMI patients: the framework requires no prior knowledge of MI and converges to final subject-specific TMP estimates after several passes of estimation with intermediate feedback; based on the primary features of the estimated spatiotemporal TMP dynamics, we provide 3-D imaging of scar tissue and quantitative evaluation of scar location and extent. Phantom experiments were performed on a computational model of realistic heart-torso geometry, considering 87 transmural infarct scars of different sizes and locations inside the myocardium, and 12 compact infarct scars (extent between 10% and 30%) at different transmural depths. Real-data experiments were carried out on BSP and magnetic resonance imaging (MRI) data from four postMI patients, validated by gold standards and existing results. This framework shows unique advantage of noninvasive, quantitative, computational imaging of subject-specific TMP dynamics and infarct mass of the 3-D myocardium, with the potential to reflect details in the spatial structure and tissue composition/heterogeneity of 3-D infarct scar.     

Frequency analysis of the electrocardiogram with maximum entropymethod for identification of patients with sustained ventriculartachycardia

Late potentials in the terminal phase of the QRS-complex during sinus rhythm have been proposed to identify a subgroup of patients with myocardial infarction at risk of ventricular tachycardia (VT). Frequency analysis of the ECG with Fourier transform (FFT) has been applied for detection of these microvolt level signals, but is limited by poor frequency resolution of short data segments and spectral leakage. We therefore developed frequency analysis using the maximum entropy method (MEM) based on an autoregressive (AR) model. Orthogonal electrocardiograms were recorded from the body surface of patients with and without VT, and healthy persons after low noise, high-gain amplification. Multiple 40 ms segments (time intervals 2 ms, AR-parameters tapered) were analyzed (spectrotemporal mapping): low-frequency components were eliminated by building difference spectra with optimal high order and fixed low order. The MEM-spectra revealed high frequency components (40-200 Hz) in the terminal phase of the QRS-complex and in the ST-section in 26/38 patients with VT, but only in 2/20 without VT and in 1/20 healthy persons (p less than 0.05). Unlike FFT, MEM allowed localization of late potentials by the analysis of short data segments. Thus, MEM offers promise for noninvasive identification of patients with sustained VT after myocardial infarction and detailed analysis of late potentials.     

A theoretical model of the high-frequency arrhythmogenic depolarization signal following myocardial infarction

Theoretical body-surface potentials were computed from single, branching and tortuous strands of Luo-Rudy dynamic model cells, representing different areas of an infarct scar. When action potential (AP) propagation either in longitudinal or transverse direction was slow (3-12 cm/s), the depolarization signals contained high-frequency (100-300 Hz) oscillations. The frequencies were related to macroscopic propagation velocity and strand architecture by simple formulas. Next, we extended a mathematical model of the QRS-complex presented in our earlier work to simulate unstable activation wavefront. It combines signals from different strands with small timing fluctuations relative to a large repetitive QRS-like waveform and can account for dynamic changes of real arrhythmogenic micropotentials. Variance spectrum of wavelet coefficients calculated from the composite QRS-complex contained the high frequencies of the individual abnormal signals. We conclude that slow AP propagation through fibrotic regions after myocardial infarction is a source of high-frequency arrhythmogenic components that increase beat-to-beat variability of the QRS, and wavelet variance parameters can be used for ventricular tachycardia risk assessment.     

Conduction velocity restitution of the human atrium--an efficient measurement protocol for clinical electrophysiological studies

Conduction velocity (CV) and CV restitution are important substrate parameters for understanding atrial arrhythmias. The aim of this work is to (i) present a simple but feasible method to measure CV restitution in-vivo using standard circular catheters, and (ii) validate its feasibility with data measured during incremental pacing. From five patients undergoing catheter ablation, we analyzed eight datasets from sinus rhythm and incremental pacing sequences. Every wavefront was measured with a circular catheter and the electrograms were analyzed with a cosine-fit method that calculated the local CV. For each pacing cycle length, the mean local CV was determined. Furthermore, changes in global CV were estimated from the time delay between pacing stimulus and wavefront arrival. Comparing local and global CV between pacing at 500 and 300 ms, we found significant changes in seven of eight pacing sequences. On average, local CV decreased by 20 ± 15% and global CV by 17 ± 13%. The method allows for in-vivo measurements of absolute CV and CV restitution during standard clinical procedures. Such data may provide valuable insights into mechanisms of atrial arrhythmias. This is important both for improving cardiac models and also for clinical applications, such as characterizing arrhythmogenic substrates during sinus rhythm.     

Detecting ventricular tachycardia and fibrillation by complexity measure

Sinus rhythm (SR), ventricular tachycardia (VT) and ventricular fibrillation (VF) belong to different nonlinear physiological processes with different complexity. In this study, we present a novel, and computationally fast method to detect VT and VF, which utilizes a complexity measure suggested by Lempel and Ziv [1]. For a specific window length (i.e., the length of data segment to be analyzed), the method first generates a 0-1 string by comparing the raw electrocardiogram (ECG) data to a selected suitable threshold. The complexity measure can be obtained from the 0-1 string only using two simple operations, comparison and accumulation. When the window length is 7 s, the detection accuracy for each of SR, VT, and VF is 100% for a test set of 204 body surface records (34 SR, 85 monomorphic VT, and 85 VF). Compared with other conventional time- and frequency-domain methods, such as rate and irregularity, VF-filter leakage, and sequential hypothesis testing, the new algorithm is simple, computationally efficient, and well suited for real-time implementation in automatic external defibrillators (AED's).     

Phase and group-delay characteristics of signal-averagedelectrocardiograms from patients with ventricular tachycardia

Fourier analysis of the signal-averaged ECG (SAECG) has previously revealed significant differences in magnitude spectra that differentiate patients with ventricular tachycardia (VT) from those without VT. To determine additional distinguishing features in the frequency domain, the authors analyzed phase spectra of SAECG's of sinus beats from 57 patients with VT, 65 without VT, and 20 normal controls. Unwrapped phase spectra from SAECG's of the entire cardiac cycle were calculated with respect to three fiducial points: onset of the P and Q waves, and the negative of the slope of the phase (group delay for frequencies in the band, which accounted for 97.5% of the energy in the vector magnitude of the Frank SAECG leads. Phase spectra of SAECG's from patients with VT differed from the non-VT patients at frequencies ⩾21 Hz (p=0.000039) for the P-wave fiducial, at frequencies ⩾60 Hz (p=0.00085) for the Q-wave fiducial, and at frequencies ⩾62 Hz (p=0.0035) for the 97.5% energy fiducial. Group delays in SAECG's from patients with and without VT differed from 10 to 26 Hz (p=0.000016) for the P-wave fiducial, and from 14 to 24 Hz (p=0.00000070) for the Q-wave fiducial. Group delays with respect to the Q-wave fiducial in the VT patients in the 14-24 Hz band were, on average, 9 ms and 5 ms longer than those of the non-VT's and normals, respectively. Thus, phase spectra of SAECG's contain previously undetected features that together with magnitude may be helpful in improving methods for stratifying the risk of VT     

Automatic quantification of changes in bone in serial MR images of joints

Recent innovations in drug therapies have made it highly desirable to obtain sensitive biomarkers of disease progression that can be used to quantify the performance of candidate disease modifying drugs. In order to measure potential image-based biomarkers of disease progression in an experimental model of rheumatoid arthritis (RA), we present two different methods to automatically quantify changes in a bone in in-vivo serial magnetic resonance (MR) images from the model. Both methods are based on rigid and nonrigid image registration to perform the analysis. The first method uses segmentation propagation to delineate a bone from the serial MR images giving a global measure of temporal changes in bone volume. The second method uses rigid body registration to determine intensity change within a bone, and then maps these into a reference coordinate system using nonrigid registration. This gives a local measure of temporal changes in bone lesion volume. We detected significant temporal changes in local bone lesion volume in five out of eight identified candidate bone lesion regions, and significant difference in local bone lesion volume between male and female subjects in three out of eight candidate bone lesion regions. But the global bone volume was found to be fluctuating over time. Finally, we compare our findings with histology of the subjects and the manual segmentation of bone lesions.     

Gait phase information provided by sensory nerve activity duringwalking: applicability as state controller feedback for FES

In this study, we extracted gait-phase information from natural sensory nerve signals of primarily cutaneous origin recorded in the forelimbs of cats during walking on a motorized treadmill. Nerve signals were recorded in seven cats using nerve cuff or patch electrodes chronically implanted on the median, ulnar, and/or radial nerves. Features in the electroneurograms that were related to paw contact and lift-off were extracted by threshold detection. For four cats, a state controller model used information from two nerves (either median and radial, or ulnar and radial) to predict the timing of palmaris longus activity during walking. When fixed thresholds were used across a variety of walking conditions, the model predicted the timing of EMG activity with a high degree of accuracy (average error = 7.8%, standard deviation = 3.0%, n = 14). When thresholds were optimized for each condition, predictions were further improved (average error = 5.5%, standard deviation = 2.3%, n = 14). The overall accuracy with which EMG timing information could be predicted using signals from two cutaneous nerves for two constant walking speeds and three treadmill inclinations for four cats suggests that natural sensory signals may be implemented as a reliable source of feedback for closed-loop control of functional electrical stimulation (FES).     

Multiresolution wavelet analysis of evoked potentials

Neurological injury, such as from cerebral hypoxia, appears to cause complex changes in the shape of evoked potential (EP) signals. To characterize such changes we analyze EP signals with the aid of scaling functions called wavelets. In particular, we consider multiresolution wavelets that are a family of orthonormal functions. In the time domain, the multiresolution wavelets analyze EP signals at coarse or successively greater levels of temporal detail. In the frequency domain, the multiresolution wavelets resolve the EP signal into independent spectral bands. In an experimental demonstration of the method, somatosensory EP signals recorded during cerebral hypoxia in anesthetized cats are analyzed. Results obtained by multiresolution wavelet analysis are compared with conventional time-domain analysis and Fourier series expansions of the same signals. Multiresolution wavelet analysis appears to be a different, sensitive way to analyze EP signal features and to follow the EP signal trends in neurologic injury. Two characteristics appear to be of diagnostic value: the detail component of the MRW displays an early and a more rapid decline in response to hypoxic injury while the coarse component displays an earlier recovery upon reoxygenation     

A mathematical model that predicts skeletal muscle force

This study demonstrates the validity of a mathematical model that predicts the force generated by rat skeletal muscles during brief subtetanic and tetanic isometric contractions. The model consists of three coupled differential equations (ODEs). The first two equations represent the calcium dynamics and the third equation represents force dynamics. The model parameters were identified from brief trains of regularly spaces pulses [constant-frequency trains (CFTs)] that produce subtetanic muscle responses. Using these parameters, the model was able to predict isometric forces from other stimulation patterns. For the gastrocnemius muscles predictions were made for responses to CFTs with interpulse intervals (IPI's) ranging from 10 to 50 ms and variable-frequency trains (VFT's), where the initial IPI=10 ms and the remaining IPIs were identical to those used for the CFTs. For the soleus muscles predictions were made for 10-100-ms CFTs. The shape of the predicted responses closely match the experimental data. Comparisons between experimental and modeled force-time integrals, peak forces, and time-to-peak also suggest excellent agreement between the model and the experiment data. Many physiological parameters predicted by the model agree with values obtained independently by others. In conclusion, the model accurately predicts isometric forces generated by rat gastrocnemius and soleus muscles produced by brief stimulation trains     

A two-stage discrimination of cardiac arrhythmias using a total least squares-based prony modeling algorithm

In this paper, we describe a new approach for the discrimination among ventricular fibrillation (VF), ventricular tachycardia (VT) and superventricular tachycardia (SVT) developed using a total least squares (TLS)-based Prony modeling algorithm. Two features, dubbed energy fractional factor (EFF) and predominant frequency (PF), are both derived from the TLS-based Prony model. In general, EFF is adopted for discriminating SVT from ventricular tachyarrhythmias (i.e., VF and VT) first, and PF is then used for further separation of VF and VT. Overall classification is achieved by performing a two-stage process to the indicators defined by EFF and PF values, respectively. Tests conducted using 91 episodes drawn from the MIT-BIH database produced optimal predictive accuracy of (SVT, VF, VT) = (95.24%, 96.00%, 97.78%). A data decimation process is also introduced in the novel method to enhance the computational efficiency, resulting in a significant reduction in the time required for generating the feature values.     

Limitations in the rapid extraction of evoked potentials using parametric modeling

The rapid extraction of variations in evoked potentials (EPs) is of great clinical importance. Parametric modeling using autoregression with an exogenous input (ARX) and robust evoked potential estimator (REPE) are commonly used methods for extracting EPs over the conventional moving time average. However, a systematic study of the efficacy of these methods, using known synthetic EPs, has not been performed. Therefore, the current study evaluates the restrictions of these methods in the presence of known and systematic variations in EP component latency and signal-to-noise ratios (SNR). In the context of rapid extraction, variations of wave V of the auditory brainstem in response to stimulus intensity were considered. While the REPE methods were better able to recover the simulated model of the EP, morphology and the latency of the ARX-estimated EPs was a closer match to the actual EP than than that of the REPE-estimated EPs. We, therefore, concluded that ARX rapid extraction would perform better with regards to the rapid tracking of latency variations. By tracking simulated and empirically induced latency variations, we conclude that rapid EP extraction using ARX modeling is only capable of extracting latency variations of an EP in relatively high SNRs and, therefore, should be used with caution in low-noise environments. In particular, it is not a suitable method for the rapid extraction of early EP components such as the auditory brainstem potential.     

Adaptive Filterng of Evoked Potentials

We present an adaptive filtering (AF) technique for rapid processing of evoked potentials (EP). The AF algorithm minimizes the mean-square error (MSE) between successive ensembles. We demonstrate theoretically that the filter output converges to the least square estimate of the underlying signal. Computer simulations with known signal and added noise show that AF produces lower MSE than ensemble averaging. We also compare results of AF to those obtained by ensemble averaging for some EP recorded from animals and humans. For very noisy EP recordings, we propose techniques that combine AF and ensemble averaging. The AF technique shows promise for requiring fewer ensembles than averaging to attain adequate signal quality.     

CPR artifact removal from human ECG using optimal multichannel filtering

The purpose of this study was to assess whether the artifacts presented by precordial compressions during cardiopulmonary resuscitation could be removed from the human electrocardiogram (ECG) using a filtering approach. This would allow analysis and defibrillator charging during ongoing precordial compressions yielding a very important clinical improvement to the treatment of cardiac arrest patients. In this investigation we started with noise-free human ECGs with ventricular fibrillation (VF) and ventricular tachycardia (VT) records. To simulate a realistic resuscitation situation, we added a weighted artifact signal to the human ECG, where the weight factor was chosen to provide the desired signal-to-noise ratio (SNR) level. As artifact signals we used ECGs recorded from animals in asystole during precordial compressions at rates 60, 90, and 120 compressions/min. The compression depth and the thorax impedance was also recorded. In a real-life situation such reference signals are available and, using an adaptive multichannel Wiener filter, we construct an estimate of the artifact signal, which subsequently can be subtracted from the noisy human ECG signal. The success of the proposed method is demonstrated through graphic examples, SNR, and rhythm classification evaluations.     

Orthonormal (Fourier and Walsh) models of time-varying evokedpotentials in neurological injury

The hypothesis that injury-related changes in evoked potential (EP) signals can be modeled by orthonormal basis functions is tested. Two models of time-varying EP signals are evaluated: the Fourier series model (FSM) and the Walsh function model (WFM). The Fourier and Walsh coefficients are estimated with the aid of an adaptive least-mean-squares (LMS) technique. Results from computer simulations illustrate how selection of model order and of the adaptation rate of the estimator affect the signal-to-noise ratio (SNR). The FSM results in a somewhat higher steady-state SNR than does the WFM; however, the WFM is less computationally complex than is the FSM. These two orthonormal functions are used to evaluate transient response to hypoxic hypoxia in anesthetized cats. Trends of the first five frequencies (Fourier) and sequencies (Walsh) show that the lower frequencies and sequencies may be sensitive indicators of hypoxic neurological injury     

Evolutionary programming CLEAN algorithm for UWB localization images of contiguous targets

In this article, a novel scattering center extractionmethod using genetic algorithm is proposed to deal with theultra-wideband (UWB) localization image, which is calledevolutionary programming (EP) CLEAN algorithm. Because ofthe UWB characters, the ideal point scattering model and EPmethod are used in the algorithm for optimizing the UWBlocalization images. After introducing the algorithm detail, theactual model is used to realize the EP CLEAN algorithm.Compared with the conventional localization imaging algorithm,this algorithm has advantages fitting the UWB characters such asaccuracy, robustness, and better resolution, which are verified bythe numerical simulations. Therefore the EP CLEAN algorithmcould improve localization image performance to expand theUWB technique application.     

Spectral analysis of periodic fluctuations in electrocardiographicrepolarization

Repolarization alternans (RPA) indicates alternate-beat fluctuations in the temporal or spatial characteristics of the electrocardiogram (ECG) STU segment which may represent dispersion in repolarization. Spectral decomposition has revealed microvolt-level RPA which has been found to correlate with ventricular tachycardia (VT) and fibrillation, and is increasingly being used for clinical risk stratification. However, while interruptions in periodicity are known to affect spectral decomposition, their quantitative impact on RPA and its clinical utility have been poorly described. The authors therefore studied the effect of variable alignment, extrasystoles, dissimilar beats and beat exclusion on RPA magnitude in simulations and on the sensitivity and specificity of RPA for VT in a pilot clinical study, RPA magnitude was exquisitely sensitive to QRS alignment such that ±1 ms random beat misalignment reduced it by 68% in simulations. Correspondingly, suboptimal QRS alignment in clinical ECG's caused the sensitivity of RPA for inducible VT to fall from 93% to as low as 63%; while JT alignment was also less effective for RPA recovery. As an experiment in minimizing morphometric irregularities in clinical ECG's, the authors found that RPA magnitude actually fell when replacing either measurably dissimilar or ectopic beats with more representative beats. In addition, inserting or deleting beats also reduced RPA magnitude in clinical sequences and simulations. These statistical analyses suggest that the precision of beat alignment and interruptions to ECG periodicity, which may occur physiologically, may greatly reduce the clinical utility of RPA for VT. Dynamic alterations in RPA in response to sequence irregularities require further study before RPA may be optimally applied to screen for ventricular arrhythmias