A method for determining high-resolution activation time delays inunipolar cardiac mapping

Presents a method for determining activation time delays in unipolar cardiac mapping data to resolutions considerably smaller than the sample interval. The method involves taking two filtered, differentiated electrograms and computing the Hilbert transform of their cross correlation, which exhibits a negative-to-positive zero crossing at the delay time between the signals. Simultaneous endocardial/epicardial recordings of sinus rhythm were made in the swine right atrium using identical, precisely superpositioned electrode arrays. Data were amplified, lowpass filtered, and digitized at 1000 Hz. A window of data was chosen around each electrogram in an endocardial/epicardial electrogram pair. The windowed electrograms were differentiated and highpass filtered, and the Hilbert transform of the cross correlation between the electrograms was computed. The activation time delay was taken to be the first negative-to-positive zero crossing. Average activation time delays (±SD) were computed for 4-s sinus rhythm recordings from each endocardial/epicardial electrode pair. For a representative site, the average transmural activation time delay was 0.71±0.06 ms (n=10 electrograms). Time delays estimated using the Hilbert transform method were compared with time delays estimated using the maximum negative slope criterion. The Hilbert transform results exhibited much smaller standard deviations, indicating that the Hilbert transform method may produce more accurate time delay estimates than the maximum negative slope method     

Spatial regularization of the electrocardiographic inverse problem and its application to endocardial mapping

Numeric regularization methods for solving the inverse problem of electrocardiography in realistic volume conductor models have been mostly limited to uniform regularization in the spatial domain. A method of spatial regularization (SR) was developed and tested in canine, where each spatial spectral component of the volume conductor model was considered separately, and a SR operator was selected based on explicit a posteriori criterion at each time instant through the heartbeat. The inverse problem was solved in the left ventricle by reconstructing endocardial surface electrograms based on cavitary electrograms measured with the use of a noncontact, multielectrode probe. The results were validated based on electrograms measured in situ at the same endocardial locations using an integrated, multielectrode basket-catheter. A probe-endocardium three-dimensional model was determined from multiplane fluoroscopic images. The boundary element method was applied to solve the boundary value problem and derive the relationship between endocardial and probe potentials. Endocardial electrograms were reconstructed during both normal and paced rhythms using SR as well as standard, uniform, zeroth-order Tikhonov (ZOT) regularization. Compared to endocardial electrograms measured by the basket, electrograms reconstructed using SR [relative error (RE) = 0.32, correlation coefficient (CC) = 0.97, activation error = 3.3 ms] were superior to electrograms reconstructed using ZOT regularization (RE = 0.59, CC = 0.79, activation error = 4.9 ms). Therefore, regularization based on spatial spectral components of the model improves the solution of the inverse problem of electrocardiography compared to uniform regularization.     

Three-dimensional imaging of ventricular activation and electrograms from intracavitary recordings

Three-dimensional (3-D) mapping of the ventricular activation is of importance to better understand the mechanisms and facilitate management of ventricular arrhythmias. The goal of this study was to develop and evaluate a 3-D cardiac electrical imaging (3DCEI) approach for imaging myocardial electrical activation from the intracavitary electrograms (EGs) and heart-torso geometry information over the 3-D volume of the heart. The 3DCEI was evaluated in a swine model undergoing intracavitary noncontact mapping (NCM). Each animal's preoperative MRI data were acquired to construct the heart-torso model. NCM was performed with the Ensite 3000 system during acute ventricular pacing. Subsequent 3DCEI analyses were performed on the measured intracavitary EGs. The estimated initial sites (ISs) were compared to the precise pacing locations, and the estimated activation sequences (ASs) and EGs were compared to those recorded by the NCM system over the endocardial surface. In total, six ventricular sites from two pigs were paced. The averaged localization error of IS was 6.7 ± 2.6 mm. The endocardial ASs and EGs as a subset of the estimated 3-D solutions were consistent with those reconstructed from the NCM system. The present results demonstrate that the intracavitary-recording-based 3DCEI approach can well localize the sites of initiation and can obtain physiologically reasonable ASs as well as EGs in an in vivo setting under control/paced conditions. This study suggests the feasibility of tomographic imaging of 3-D ventricular activation and 3-D EGs from intracavitary recordings.     

Noninvasive mapping of transmural potentials during activation in Swine hearts from body surface electrocardiograms

The three-dimensional cardiac electrical imaging (3DCEI) technique was previously developed to estimate the initiation site(s) of cardiac activation and activation sequence from the noninvasively measured body surface potential maps (BSPMs). The aim of this study was to develop and evaluate the capability of 3DCEI in mapping the transmural distribution of extracellular potentials and localizing initiation sites of ventricular activation in an in vivo animal model. A control swine model (n = 10) was employed in this study. The heart-torso volume conductor model and the excitable heart model were constructed based on each animal's preoperative MR images and a priori known physiological knowledge. Body surface potential mapping and intracavitary noncontact mapping (NCM) were simultaneously conducted during acute ventricular pacing. The 3DCEI analysis was then applied on the recorded BSPMs. The estimated initiation sites were compared to the precise pacing sites; as a subset of the mapped transmural potentials by 3DCEI, the electrograms on the left ventricular endocardium were compared to the corresponding output of the NCM system. Over the 16 LV and 48 RV pacing studies, the averaged localization error was 6.1±2.3 mm, and the averaged correlation coefficient between the estimated endocardial electrograms by 3DCEI and from the NCM system was 0.62±0.09. The results demonstrate that the 3DCEI approach can well localize the sites of initiation of ectopic beats and can obtain physiologically reasonable transmural potentials in an in vivo setting during focal ectopic beats. This study suggests the feasibility of tomographic mapping of 3D ventricular electrograms from the body surface recordings.     

Solving the inverse problem of electrocardiography using a Duncan and Horn formulation of the Kalman filter

Numeric regularization methods most often used to solve the ill-posed inverse problem of electrocardiography are spatial and ignore the temporal nature of the problem. In this paper, a Kalman filter reformulation incorporated temporal information to regularize the inverse problem, and was applied to reconstruct left ventricular endocardial electrograms based on cavitary electrograms measured by a noncontact, multielectrode probe. These results were validated against in situ electrograms measured with an integrated, multielectrode basket-catheter. A three-dimensional, probe-endocardium model was determined from multiplane fluoroscopic images. The boundary element method was applied to solve the boundary value problem and determine a linear relationship between endocardial and probe potentials. The Duncan and Horn formulation of the Kalman filter was employed and was compared to the commonly used zero- and first-order Tikhonov spatial regularization as well as the Twomey temporal regularization method. Endocardial electrograms were reconstructed during both sinus and paced rhythms. The Paige and Saunders solution of the Duncan and Horn formulation reconstructed endocardial electrograms at an amplitude relative error of 13% (potential amplitude) which was superior to solutions obtained with zero-order Tikhonov (relative error, 31%), first-order Tikhonov (relative error, 19%), and Twomey regularization (relative error, 44%). Likewise, activation time error in the inverse solution using the Duncan and Horn formulation (2.9 ms) was smaller than that of zero-order Tikhonov (4.8 ms), first-order Tikhonov (5.4 ms), and Twomey regularization (5.8 ms). Therefore, temporal regularization based on the Duncan and Horn formulation of the Kalman filter improves the solution of the inverse problem of electrocardiography.     

Model-based imaging of cardiac electrical excitation in humans

Activation time (AT) imaging from electrocardiographic (ECG) mapping data has been developing for several years. By coupling ECG mapping and three-dimensional (3-D) + time anatomical data, the electrical excitation sequence can be imaged completely noninvasively in the human heart. In this paper, a bidomain theory-based surface heart model AT imaging approach was applied to single-beat data of atrial and ventricular depolarization in two patients with structurally normal hearts. In both patients, the AT map was reconstructed from sinus and paced rhythm data. Pacing sites were the apex of the right ventricle and the coronary sinus (CS) ostium. For CS pacing, the reconstructed AT pattern on the endocardium of the right atrium was compared with the CARTO map in both patients. The localization errors of the origins of the initial endocardial breakthroughs were determined to be 6 and 12 mm. The sites of early activation and the areas with late activation were estimated with sufficient accuracy. The reconstructed sinus rhythm sequence was in good qualitative agreement with the pattern previously published for the isolated Langendorff-perfused human heart.     

An interactive computer system for guiding the surgical treatment of life-threatening ventricular tachycardias

Ventricular tachycardias that are medically refractory can be treated surgically by resection of the area that generates the arrhythmia. The origin of a tachycardia can be determined during operation by recording endocardial electrograms. Tachycardias occurring spontaneously, or induced during the operative procedure, often last for only a few cycles and require the simultaneous recording of endocardial signals from numerous sites (mapping). The conventional technique of sequential mapping by using a roving probe cannot be applied in such cases. We, therefore, developed a flexible computerized system which enables us to record simultaneously from a multiple of 16 electrode terminals. The basic concept can accommodate 64 channels. A multielectrode array consisting of an inflatable balloon, uniformly covered with electrode terminals, is used to map the whole endocardial surface simultaneously. A rectangular grid covered with electrode terminals is used for refined resolution. Optically isolated amplifiers are used, providing ohmic isolation between the patient and the support electronics. After amplification and filtering, the endocardial signals are collected in groups of 16 and transferred to data acquisition units. Each unit contains a 16-channel multiplexing A/D converter and a 16K circular buffer. An LSI-11 based minicomputer controls the acquisition units and accounts for the transition of data to disk. Preselection of data is carried out during acquisition in order to reduce operation time. After acquisition, hard copies of the signals are made on a 16-channel chart recorder.     

The Effects of Distant Cardiac Electrical Events on Local Activation in Unipolar Epicardial Electrograms

The effect of potentials generated during depolarization of the left ventricle on epicardial unipolar electrograms recorded from the right ventricle was studied using the right ventricular isolation procedure. This surgical technique disrupts electrical conduction and prevents activation wavefronts from propagating between the ventricles. Following the procedure, the ventricles were paced asynchronously with the left ventricle paced 100 ms before the right ventricle to geparate the electrogram into its local (due to depolarization of the right ventricle) and distant (due to depolarization of the left ventricle) components or with an interval of 20 ms or less between pacing the ventricles to mimic electrograms resulting from normal (synchronous) activration. Electrical activity in the left ventricle significantly affected the magnitude of the slope of the most rapid deflection and the timing of the maximum and minimum potentials of right ventricular unipolar electrograms. However, distant activity did not significantly alter the timing of the fastest 1 ms downstroke. No electrograms of distant components had negative slopes with magnitudes greater than 1.3 mV/ms, nor did any slopes of electrograms containing only local components have magnitudes less than 1.5 mV/ms. Simulated electrograms, calculated from the local and distant components, correlated well (r = 0.83 to 1.00, N = 48) with electrograms recorded during synchronous pacing.     

Multiway sequential hypothesis testing for tachyarrhythmiadiscrimination

A multiway sequential hypothesis testing (M-SHT) algorithm is proposed for simultaneous discrimination of cardiac tachyarrhythmias-supraventricular tachycardia (SVT) and ventricular tachycardia (VT)-from normal sinus rhythm (NSR). The M-SHT algorithm calculates a likelihood function from atrio-ventricular delay measurements, and compares this function with thresholds derived from specified error probabilities for the arrhythmias to be discriminated. Performance of this algorithm was evaluated on dual channel endocardial electrograms recorded in the cardiac electrophysiology laboratory. Two databases were developed, one for development of the algorithm and another for evaluation. The M-SHT algorithm accurately classified 26 out of 28 NSR (2 misclassified as SVT), 31 out of 31 cases of SVT, and 41 out of 43 VT (2 misclassified as NSR). The average length of time taken for classification of the three rhythms was: 3.6 s for NSR, 5.0 s for SVT, and 1.6 s for VT. Unique features of this algorithm are that acceptable error rates for each arrhythmia are independently specified and accuracy can be traded off for a faster detection time, and vice versa     

Automatic detection of conduction block based on time-frequencyanalysis of unipolar electrograms

It is commonly thought that lethal tachyarrhythmias, such as ventricular fibrillation (VF), are perpetuated by functional reentry, which occurs when an activation wave blocks and rotates around tissue that is excitable (i.e., functional block). Electrograms recorded near these regions typically contain two sequential deflections representing activation on either side of the block. By detecting these "double potentials," we hypothesize that functional block can be detected by a single electrode. METHODS: Unipolar electrograms were recorded from a 24 x 21 mapping array on the intact ventricular epicardium of five pigs during electrically-induced VF. The short time Fourier Transform (STFT) of each electrogram was analyzed to identify double potentials. To evaluate the performance of the STFT algorithm, conduction block was located in activation maps using a minimum conduction velocity criterion (10 cm/s) and then compared to the results of the STFT algorithm. RESULTS: The STFT algorithm detected conduction block with a sensitivity of 0.74 +/- 0.12 and a specificity of 0.99 +/- 0.00. CONCLUSION: We have developed an automated algorithm that can detect functional block during VF from a single electrode recording. Possible applications include fast, objective identification of block in mapping data and realtime localization of reentrant substrates using mapping catheters.     

On the Possibility of Directly Relating the Pattern of Ventricular Surface Activation to the Pattern of Body Surface Potential Distribution

A system of three 320-element spheres was employed to represent the endocardial and epicardial surfaces of the left ventricle and the body surface. The two inner (heart) spheres were considered electrogenic, and each active subunit was given an onset time and a monophasic action potential; these subunits were treated as source dipoles for successive instants in time. The potential distribution at any instant resulting on the outer (torso) surface was calculated from adding together the corresponding proportionate effects of all active subunits, each treated as dipolar sources. This result was compared to multipolar reduction of simultaneous endocardial and epicardial action potential patterns which, when combined, gave a net multipolar generator content enabling outer pattern approximation. The identity between the patterns of torso surface potential, systematically calculated from multiple dipoles, and those produced from the multipolar reduction provided three insights: 1) the whole surface treatment of the multipolar method is faster, 2) both show an offset term related to the monophasic nature of the sources and similar to that found in live data, and 3) such a model may provide a vehicle for experimentally testing the contribution of intramural sources to body surface potential maps.     

A comparison of noninvasive reconstruction of epicardial versus transmembrane potentials in consideration of the null space

We compare two source formulations for the electrocardiographic forward problem in consideration of their implications for regularizing the ill-posed inverse problem. The established epicardial potential source model is compared with a bidomain-theory-based transmembrane potential source formulation. The epicardial source approach is extended to the whole heart surface including the endocardial surfaces. We introduce the concept of the numerical null and signal space to draw attention to the problems associated with the nonuniqueness of the inverse solution and show that reconstruction of null-space components is an important issue for physiologically meaningful inverse solutions. Both formulations were tested with simulated data generated with an anisotropic heart model and with clinically measured data of two patients. A linear and a recently proposed quasi-linear inverse algorithm were applied for reconstructions of the epicardial and transmembrane potential, respectively. A direct comparison of both formulations was performed in terms of computed activation times. We found the transmembrane potential-based formulation is a more promising source formulation as stronger regularization by incorporation of biophysical a priori information is permitted.     

Four digital algorithms for activation detection from unipolarepicardial electrograms

The reproducibility of activation detection by each of four algorithms used to calculate maximum derivatives was tested on two sequential paced beats of right ventricular unipolar epicardial electrograms which represented either local activation of the right ventricle alone or synchronous activation of both ventricles. The methods were evaluated by comparing the shape of the two beats aligned on their selected activation times, i.e., the time at which the maximum negative deflection occurred, the differences in activation intervals for the two beats, and the effect on the activation time of superimposing distant events on local activation. The 17-point second-order data fit algorithm performed slightly better than the first-order difference, three-point Lagrange derivative, and five-point second-order data fit algorithms except that activation time selection by the 17-point technique was slightly, but significantly, delayed by the superposition of distant potentials. The 17-point second-order data fit technique is therefore recommended for use in detecting activation unless computation time is a major consideration. In that case, the five-point second-order data fit technique, which uses only four data values for each computation, can be used with only slight decreases in accuracy.     

Ventricular Intramyocardial Electrograms and Their Expected Potential for Cardiac Risk Surveillance, Telemonitoring, and Therapy Management

Ventricular intramyocardial electrograms are recorded with electrodes directly from the heart either in intraventricular or epimyocardial position and may be acquired either from the spontaneously beating or from the paced heart. The morphology of these signals differs significantly from that of body surface ECG recordings. Although the morphology shows general characteristics, it additionally depends on different individual impacts. This problem of individual evaluation is briefly discussed. As an appropriate methodology for its solution, personalized referencing based on similarity averaging has been employed. A more general approach may be model-based signal interpretation, which is still under investigation. The preliminary results reveal a promising potential of intramyocardial electrograms for cardiac risk surveillance, e.g., for arrhythmia detection, recognition of rejection events in transplanted hearts, and assessment of hemodynamic performance. Employing implants with telemetric capabilities may render possible permanent and even continuous cardiac telemonitoring. Furthermore, the signals can be utilized for supporting therapy management, e.g., in patients with different kinds of cardiomyopathies. This paper shall demonstrate some preliminary results and discuss the expected potential.      

Enhancement of contrast echocardiography by image variabilityanalysis

BACKGROUND: Although there have been recent advances in echocardiography, many studies remain suboptimal due to poor image quality and unclear blood-myocardium border. We developed a novel image processing technique, cardiac variability imaging (CVI), based on the variance of pixel intensity values during passage of ultrasound microbubble contrast into the left ventricle chamber, with the aim of enhancing endocardial border delineation and image quality. METHODS AND RESULTS: CVI analysis was performed on simulated data to test and verify the mechanism of image enhancement. Then CVI analysis was applied to echocardiographic images obtained in two different clinical studies, and still images were interpreted by expert reviewers. In the first study (N = 15), using contrast agent EchoGen, the number of observable wall segments in end-diastolic images, for example, was significantly increased by CVI (4.93) as compared to precontrast (3.28) and contrast images (3.36), P < 0.001 for both comparisons to CVI. In the second study (N = 8), using contrast agent Optison, interobserver variability of manually traced end-diastolic volumes was significantly decreased using CVI (22.3 ml) as compared to precontrast (63.4) and contrast images (49.0), P < 0.01 for both comparisons to CVI. CONCLUSION: CVI can substantially enhance endocardial border delineation and improve echocardiographic image quality and image interpretation.     

Automatic segmentation of echocardiographic sequences by active appearance motion models

A novel extension of active appearance models (AAMs) for automated border detection in echocardiographic image sequences is reported. The active appearance motion model (AAMM) technique allows fully automated robust and time-continuous delineation of left ventricular (LV) endocardial contours over the full heart cycle with good results. Nonlinear intensity normalization was developed and employed to accommodate ultrasound-specific intensity distributions. The method was trained and tested on 16-frame phase-normalized transthoracic four-chamber sequences of 129 unselected infarct patients, split randomly into a training set (n = 65) and a test set (n = 64). Borders were compared to expert drawn endocardial contours. On the test set, fully automated AAMM performed well in 97% of the cases (average distance between manual and automatic landmark points was 3.3 mm, comparable to human interobserver variabilities). The ultrasound-specific intensity normalization proved to be of great value for good results in echocardiograms. The AAMM was significantly more accurate than an equivalent set of two-dimensional AAMs.     

A Waveguide Band-Stop Filter as an Image Rejection Filter for Measurement of Stratospheric Ozone

In this report we present the performance and test observation results of a waveguide band-stop filter (BSF) as an image rejection filter for the measurement of stratospheric ozone. By using the waveguide BSF, we are able to adopt a very simple optical system and achieve a good image rejection ratio. Additionally, we are able to observe in both single sideband (SSB) mode and double sideband (DSB) mode by only changing the local oscillator (LO) frequency. We have installed the waveguide BSF into an atmospheric ozone-measuring system using a superconductive (SIS) receiver and have successfully observed an ozone spectrum at 110 GHz in SSB and DSB mode. The receiver noise temperature (SSB) and the image rejection ratio at 110 GHz are about 60 K and more than 30 dB, respectively. Because of the IF power ripple, the waveguide BSF cannot be used with a wide-band spectrometer. However, it is quite practical for narrow-band observation.     

Efficient electrode spacing for examining spatial organizationduring ventricular fibrillation

Spatial organization has been observed during episodes of ventricular fibrillation (VF) by recording epicardial unipolar electrograms on a grid of electrodes. In such studies, the choice of spacing between electrodes is an important decision, affecting the resolution and the size of the domain to be studied. A basic tenet of sampling theory, the Nyquist criterion, states that an electrode spacing smaller than half the smallest significant wavelength is required to capture the important details of a spatially sampled process. The authors suggest a method to choose a practical interelectrode spacing by examining wavenumber power spectra of high-resolution VF data recorded from a square 11×11 array of electrodes spaced 0.28 mm apart. The plaque was sutured on the epicardium near the left ventricular apex in 7 anesthetized pigs. VF was induced with AC simulation. Unipolar extracellular electrograms were simultaneously recorded from each channel for 2 s after the induction of VF. Each signal was sampled in time at 1000 Hz. Wavenumber power spectra were calculated for 100 ms segments using the zero-delay wavenumber spectrum method, for a total of 140 power spectra. All spectra had dominant peaks at the origin and fell off rapidly with increasing wavenumber (decreasing wavelength). In all the spectra, every wavelength shorter than 1.4 mm contributed insignificant power. Furthermore, in 134 of 140 spectra (96%), insignificant power levels were associated with every wavelength shorter than 2.8 mm. These results suggest that, for unipolar extracellular electrodes, an intersensor spacing on the order of 1 mm is appropriate to study organization during early VF     

Imaging myocardial strain

Measuring the local mechanical activity of the heart has lagged behind the measurement of electrical activity due to a lack of measurement tools. Myocardial wall motion abnormalities have been studied for years in the context of regional ischemia. Implanted beads and screws have been used to measure the mechanical activity of the heart in a few isolated regions. Over the past decade, precise and accurate methods for measuring local three-dimensional (3-D) myocardial motion with magnetic resonance imaging (MRI) have been developed using presaturation tagging patterns, velocity encoded phase maps, and displacement encoded phase maps. Concurrently, the quality of cardiac MRI images improved greatly with the use of customized receiver coils and the speed of acquisition has increased dramatically with the advent of undersampling techniques and new generations of MR machines with faster switching gradient coils. The use of these cardiac MRI techniques to produce an image of the local deformation of the heart in the form of a myocardial strain image is described. Using these images, the “mechanical activation” of the heart are defined, that is, the time of onset of contraction. A map of the mechanical activation over the heart is a direct analogy to an electrical activation map of the heart