Beatriz Chicote

Circulation detection using the electrocardiogram and the thoracic impedance acquired by defibrillation pads

AIM To develop and evaluate a method to detect circulation in the presence of organized rhythms (ORs) during resuscitation using signals acquired by defibrillation pads. METHODS Segments containing electrocardiogram (ECG) and thoracic impedance (TI) signals free of artifacts were used. The ECG corresponded to ORs classified as pulseless electrical activity (PEA) or pulse-generating rhythm (PR). A first dataset containing 1091 segments was split into training and test sets to develop and validate the circulation detector. The method processed ECG and TI to obtain the impedance circulation component (ICC). Morphological features were extracted from ECG and ICC, and combined into a classifier to discriminate between PEA and PR. The performance of the method was evaluated in terms of sensitivity (PR) and specificity (PEA). A second dataset (86 segments from different patients) was used to assess two application of the method: confirmation of arrest by recognizing absence of circulation during ORs and detection of return of spontaneous circulation (ROSC) during resuscitation. In both cases, time to confirmation of arrest/ROSC was determined. RESULTS The method showed a sensitivity/specificity of 92.1%/90.3% and 92.2%/91.9% for training and test sets respectively. The method confirmed cardiac arrest with a specificity of 93.3% with a median delay of 0s after the first OR annotation. ROSC was detected with a sensitivity of 94.4% with a median delay of 57s from ROSC onset. CONCLUSION The method showed good performance, and can be reliably used to distinguish perfusing from non-perfusing ORs.

A method to measure ventilation rate during cardiopulmonary resuscitation using the capnogram

The survival rate in cardiac arrest is associated to the quality of the chest compressions (CCs) and ventilations provided during cardiopulmonary resuscitation (CPR). Hyperventilation remains common whenever ventilation is manual during resuscitation from cardiac arrest. The capnogram is used to monitor respiration and ventilation rates. During CPR chest compressions induce artefacts in the capnogram signal that challenge the detection of ventilations. The evaluation of ventilation detectors during CCs has not been well characterized. In this study an algorithm for ventilation rate monitoring and hyperventilation detection was developed. The processing method consists of detecting transitions in the first difference of the signal, and applying feature based classification to identify every ventilation. The instantaneous rate and hyperventilation minutes were then computed. A set of 20 out-of-hospital episodes, totalling 50864 s (86.6% during CCs) and 6305 ventilations was used to define and evaluate the algorithm. The algorithm had a sensitivity/ positive predictive value of 96.9%/96.2% respectively for the ventilation detection (96.7%/95.8% during ongoing CCs), 98.7%/98.7% for the hyperventilation detection, and a mean error of 0.4 (0.8) min-1 for the instantaneous ventilation rate.

Alternatives to Estimate the Compression Depth from the Acceleration Signal during Cardiopulmonary Resuscitation

Feedback devices improve the quality of chest compressions during cardiopulmonary resuscitation. Most devices use accelerometers to estimate the chest displacement using double integration and additional reference signals to ensure the stability of the process. We described and evaluated three novel methods for computing the compression depth using solely the acceleration signal. The BPF method approximates the integration using a band-pass filter. The ZCV method computes the velocity and calculates the depth from the zero-crossing instants of each compression cycle. The SAA method computes the depth from the spectral analysis of the acceleration signal. We gathered twelve 10-min records in which chest compressions were provided on a manikin equipped with a displacement sensor. A tri -axial accelerometer was placed beneath the rescuers hands. The median (IQR) unsigned error in mm between the computed depth and the reference was 4.0 (2.1-6.2), 3.9 (2.0-6.2), and 1.2 (0.6-2.1) for each method, respectively. Compression depth can be accurately estimated from chest acceleration. The spectral analysis method provided the best global performance. This alternative could be implemented for real time feedback during CPR.

Fuzzy and Sample Entropies as Predictors of Patient Survival Using Short Ventricular Fibrillation Recordings during out of Hospital Cardiac Arrest

Optimal defibrillation timing guided by ventricular fibrillation (VF) waveform analysis would contribute to improved survival of out-of-hospital cardiac arrest (OHCA) patients by minimizing myocardial damage caused by futile defibrillation shocks and minimizing interruptions to cardiopulmonary resuscitation. Recently, fuzzy entropy (FuzzyEn) tailored to jointly measure VF amplitude and regularity has been shown to be an efficient defibrillation success predictor. In this study, 734 shocks from 296 OHCA patients (50 survivors) were analyzed, and the embedding dimension (m) and matching tolerance (r) for FuzzyEn and sample entropy (SampEn) were adjusted to predict defibrillation success and patient survival. Entropies were significantly larger in successful shocks and in survivors, and when compared to the available methods, FuzzyEn presented the best prediction results, marginally outperforming SampEn. The sensitivity and specificity of FuzzyEn were 83.3% and 76.7% when predicting defibrillation success, and 83.7% and 73.5% for patient survival. Sensitivities and specificities were two points above those of the best available methods, and the prediction accuracy was kept even for VF intervals as short as 2s. These results suggest that FuzzyEn and SampEn may be promising tools for optimizing the defibrillation time and predicting patient survival in OHCA patients presenting VF.

Sample entropy as a shock outcome predictor during basis life support

Optimizing defibrillation times may improve survival from ventricular fibrillation (VF) cardiac arrest. VF waveform analysis is one of the best non-invasive decision tools for shock outcome prediction. This study introduces a VF-waveform feature based on the computation of the sample entropy (SmpEnt) for shock outcome prediction. A database of 255 shocks were analyzed, using a 5 s preshock ECG segment. 14 classical VF waveform features measuring amplitude, slope, complexity and spectral characteristics were computed in addition to SmpEnt. An optimal detector of successful shocks was designed for each feature minimizing the Balanced Error Rate. Finally, the minimum pres hock segment duration assuring an accurate shock outcome prediction was determined for SmpEnt. SmpEnt is an improved shock outcome predictor, even for VF-segments as short as 1.5-s, and it could be used as a decision support tool to guide optimal timing for defibrillation.

Feasibility of compression depth estimation from the acceleration signal during cardiopulmonary resuscitation in long-distance trains

Early cardiopulmonary resuscitation (CPR) and early defibrillation improve survival from out-of-hospital cardiac arrest. Long distance trains are increasingly being equipped with defibrillators. CPR feedback devices help rescuers to deliver chest compressions with an adequate depth. Most of them are based on accelerometers placed beneath the rescuers hands. However, in a moving train the measured acceleration is a combination of the acceleration of the chest and that of the train. We wanted to evaluate the accuracy of two accelerometer-based systems in this scenario. Chest compressions were delivered on a resuscitation manikin during the Zaragoza-Bilbao (Spain) ALVIA train route. A tri-axial accelerometer was placed between the manikin s chest and the rescuer s hands. We acquired 3min records between consecutive stations with compressions delivered at target depths of 35 and 50 mm. Records corresponding to intervals with diferent average train velocities were selected. We applied a time-domain (TD) method and a frequency-domain method (FD) to the acceleration records for estimating the compression depth. We analysed the implications of using a single axis (al) or composing the three axis (a3) of the acceleration. The median (IQR) unsigned error in mm was 6.4 (3.710.1), 5.9 (2.9-10.1), 1.8 (0.8-3.1), and 2.0 (1.0-3.6), for TDa1, TDa3, FDa1, and FDa3, respectively. Chest compression depth could be accurately estimated from the spectral analysis of the acceleration in a moving train. The accuracy of the time-domain method was severely compromised, with median errors above 10% of the target depth.