Antonio Cebrián

The role of independent component analysis in the signal processing of ECG recordings.

Abstract Independent component analysis (ICA) is an emerging technique for multidimensional signal processing. In recent years, these techniques have been proposed for solving a large number of biomedical applications. This work reviews current knowledge on ICA in electrocardiographic (ECG) analysis. The benefits that ICA can bring to clinical practice are illustrated with four relevant clinical applications: foetal ECG extraction from maternal ECG recordings, analysis of atrial fibrillation, ECG denoising and removal of pacemaker artefacts.

Efecto del estiramiento miocárdico sobre las frecuencias de activación determinadas mediante análisis espectral durante la fibrilación ventricular

Introduction and objectives. The aim of this study was to analyze the effects of myocardial stretching on excitation frequencies, as determined by spectral analysis, during ventricular fibrillation. Methods. In 12 isolated rabbit heart preparations, ventricular activation during ventricular fibrillation was recorded with multiple electrodes. Recordings were obtained before, during and after ventricular dilatation produced with an intraventricular balloon. The dominant frequency of the signals obtained with each of the electrodes was determined by spectral analysis. Results. During the control phase, the mean, minimum and maximum dominant frequencies were, respectively, 14.3 ± 1.7, 12.5 ± 1.7, and 16.2 ± 1.4 Hz, and the average difference between the maximum and minimum frequencies was 3.6 ± 2.1 Hz. This difference was over 4 Hz in four cases, and in no case did it exceed 8 Hz. During ventricular stretching, the mean dominant frequency increased significantly (21.1 ± 6.1 Hz; p < 0.0001), as did the minimum values (14 ± 2.6 Hz; p < 0.05) and especially the maximum values (26.6 ± 7.7 Hz; p < 0.0001). The difference between the maximum and minimum frequencies (12.6 ± 6.4 Hz; p < 0.001) was over 4 Hz in all cases except one, and over 8 Hz in 9 cases. The maximum values were distributed heterogeneously during ventricular stretching. Upon suppressing ventricular stretching, the dominant frequency did not differ from controls. Conclusions. Myocardial frequency maps during ventricular fibrillation show limited variations in the dominant frequency of the signals recorded in the lateral wall of the left ventricle. During stretching, the patterns were heterogeneous, due mainly to the marked increase in the maximum dominant frequency. In the experimental model used, the effects of stretching remitted after suppressing ventricular dilatation.

Optimization of three morphologic algorithms for arrhythmia discrimination in implantable cardioverter defibrillators

Discrimination between ventricular tachycardia (VT) and supraventricular tachycardia (SVT) in implantable cardioverter defibrillators (ICDs) is still an unsolved task due to the low specificity of traditional techniques based in rate, stability and onset. Several morphological published algorithms enhance VT vs. SVT discrimination by increasing algorithm complexity. Three morphological published algorithms with increasing complexity have been selected: time domain (complex peak area comparison), simplified wavelet and frequency domain (Fourier complex power spectra analysis and neural network) algorithms. All them have been reconstructed from published information and programmed in MATLAB. The algorithms has been optimized in order to obtain an improved classification and to work in a 16-bit microcontroller platform (Texas Instruments MSP430 microcontroller). A final test of the optimized algorithms has been accomplished using a classified unipolar and bipolar electrogram (EGM) database. The configurable parameters of the algorithms have been adjusted in order to maximize sensitivity (SE), specificity (SP) and accuracy (AC)

Implantable cardioverter defibrillator algorithms: status review in terms of computational cost.

Abstract In recent decades, implantable cardioverter defibrillators (ICDs) have improved substantially, becoming the treatment of choice for patients at high risk of life-threatening arrhythmias. Nevertheless, inappropriate shock therapy for non-ventricular arrhythmias is still a problem. Extending the ICD battery lifetime demands very low power consumption, which is obtained at very low microprocessor clock frequencies. Currently, some high-performance algorithms remain beyond the computational capabilities of ICDs. Future ICDs with higher computing power will permit the implementation of computationally intensive algorithms, enhancing the discrimination performance and preventing inappropriate shock therapies. An ICD algorithm status review is presented from the point of view of signal processing techniques and their computational costs. Several examples of discrimination algorithms with increasing computational cost are analyzed. Whereas some of them are already used in commercial ICDs, other algorithms cannot be implemented yet in current ICDs. A solution based on dynamic adaptation of microprocessor power consumption to meet algorithm computational requirements is proposed. This solution allows implementation of complex discrimination algorithms in ICDs without significantly increasing the power consumption.

QT Interval Heterogeneities Induced Through Local Epicardial Warming/Cooling. An Experimental Study

INTRODUCTION AND OBJECTIVES Abnormal QT interval durations and dispersions have been associated with increased risk of ventricular arrhythmias. The present study examines the possible arrhythmogenic effect of inducing QT interval variations through local epicardial cooling and warming. METHODS In 10 isolated rabbit hearts, the temperatures of epicardial regions of the left ventricle were modified in a stepwise manner (from 22°C to 42°C) with simultaneous electrogram recording in these regions and in others of the same ventricle. QT and activation-recovery intervals were determined during sinus rhythm, whereas conduction velocity and ventricular arrhythmia induction were determined during programmed stimulation. RESULTS In the area modified from baseline temperature (37°C), the QT (standard deviation) was prolonged with maximum hypothermia (195 [47] vs 149 [12] ms; P<.05) and shortened with hyperthermia (143 [18] vs 152 [27] ms; P<.05). The same behavior was displayed for the activation-recovery interval. The conduction velocity decreased with hypothermia and increased with hyperthermia. No changes were seen in the other unmodified area. Repetitive responses were seen in 5 experiments, but no relationship was found between their occurrence and hypothermia or hyperthermia (P>.34). CONCLUSIONS In the experimental model employed, local variations in the epicardial temperature modulate the QT interval, activation-recovery interval, and conduction velocity. Induction of heterogeneities did not promote ventricular arrhythmia occurrence.

Epicardial-limited electrophysiological heterogeneities do not facilitate ventricular arrhythmia induction. An experimental study

The electrophysiological heterogeneities of the myocardium are associated with vulnerability to arrhythmias. This study presents an experimental heterogeneity model based on local epicardial cooling/warming. The ventricular activation-recovery interval (ARl), conduction velocity (CV) and arrhythmogenic response to electrical pacing were determined. Electrical mapping was carried out on isolated rabbit hearts (n=8), using a specific electrode device for epicardial temperature control. With respect to baseline, ARl in the modified zone was prolonged (137 ± 22 ms vs 111 ± 13 ms, p<;0.05) under maximum hypothermia (22.3 ± 0.6 °C vs 36.7 ± 0.8 DC), and was shortened (98 ± 13 ms vs 107 ± 16 ms, p<;0.05) under conditions of hyperthermia (41.8 ± 0.3 °C vs 37.3 ± 0.4 DC). CV decreased (70 ± 17 cm/s vs 76 ± 17 cm/s, p<;0.05) under hypothermia and increased (79 ± 20 cm/s vs 75 ± 21 cm/s, p<;0.05) under hyperthermia. There were no changes in the unmodified zone. Repetitive responses were observed in four hearts, with no dependency between the appearance of responses and the induced modifications. Thermally induced dispersion of ARl and CV did not favor the induction of ventricular arrhythmias, probably because only a limited zone of the ventricular epicardium was affected.