Unai Ayala

Suppression of the cardiopulmonary resuscitation artefacts using the instantaneous chest compression rate extracted from the thoracic impedance

AIM To demonstrate that the instantaneous chest compression rate can be accurately estimated from the transthoracic impedance (TTI), and that this estimated rate can be used in a method to suppress cardiopulmonary resuscitation (CPR) artefacts. METHODS A database of 372 records, 87 shockable and 285 non-shockable, from out-of-hospital cardiac arrest episodes, corrupted by CPR artefacts, was analysed. Each record contained the ECG and TTI obtained from the defibrillation pads and the compression depth (CD) obtained from a sternal CPR pad. The chest compression rates estimated using TTI and CD were compared. The CPR artefacts were then filtered using the instantaneous chest compression rates estimated from the TTI or CD signals. The filtering results were assessed in terms of the sensitivity and specificity of the shock advice algorithm of a commercial automated external defibrillator. RESULTS The correlation between the mean chest compression rates estimated using TTI or CD was r=0.98 (95% confidence interval, 0.97-0.98). The sensitivity and specificity after filtering using CD were 95.4% (88.4-98.6%) and 87.0% (82.6-90.5%), respectively. The sensitivity and specificity after filtering using TTI were 95.4% (88.4-98.6%) and 86.3% (81.8-89.9%), respectively. CONCLUSIONS The instantaneous chest compression rate can be accurately estimated from TTI. The sensitivity and specificity after filtering are similar to those obtained using the CD signal. Our CPR suppression method based exclusively on signals acquired through the defibrillation pads is as accurate as methods based on signals obtained from CPR feedback devices.

A high-temporal resolution algorithm to discriminate shockable from nonshockable rhythms in adults and children

AIM To design the core algorithm of a high-temporal resolution rhythm analysis algorithm for automated external defibrillators (AEDs) valid for adults and children. Records from adult and paediatric patients were used all together to optimize and test the performance of the algorithm. METHODS A total of 574 shockable and 1126 nonshockable records from 1379 adult patients, and 57 shockable and 503 nonshockable records from 377 children aged between 1 and 8 years were used. The records were split into two groups for development and testing. The core algorithm analyses ECG segments of 3.2s duration and classifies the segments as nonshockable or likely shockable combining a time, slope and frequency domain analysis to detect normally conducted QRS complexes. RESULTS The algorithm correctly identified 98% of nonshockable segments, 97.5% in adults and 98.4% in children, and identified 99.5% of shockable segments as likely shockable, 100% in adults and 96% in children. When likely shockable segments were further analysed in terms of regularity, spectral content and heart rate to form a complete rhythm analysis algorithm the overall specificity increased to 99.6% and the sensitivity was 99.1%. CONCLUSION Paediatric and adult rhythms can be accurately diagnosed using 3.2s ECG segments. A single algorithm safe for children and adults can simplify AED use, and its high temporal resolution shortens pre-shock pauses which may contribute to improve resuscitation outcome.

Rhythm Analysis during Cardiopulmonary Resuscitation: Past, Present, and Future

Survival from out-of-hospital cardiac arrest depends largely on two factors: early cardiopulmonary resuscitation (CPR) and early defibrillation. CPR must be interrupted for a reliable automated rhythm analysis because chest compressions induce artifacts in the ECG. Unfortunately, interrupting CPR adversely affects survival. In the last twenty years, research has been focused on designing methods for analysis of ECG during chest compressions. Most approaches are based either on adaptive filters to remove the CPR artifact or on robust algorithms which directly diagnose the corrupted ECG. In general, all the methods report low specificity values when tested on short ECG segments, but how to evaluate the real impact on CPR delivery of continuous rhythm analysis during CPR is still unknown. Recently, researchers have proposed a new methodology to measure this impact. Moreover, new strategies for fast rhythm analysis during ventilation pauses or high-specificity algorithms have been reported. Our objective is to present a thorough review of the field as the starting point for these late developments and to underline the open questions and future lines of research to be explored in the following years.

Automatic detection of chest compressions for the assessment of CPR-quality parameters

AIM Accurate chest compression detection is key to evaluate cardiopulmonary resuscitation (CPR) quality. Two automatic compression detectors were developed, for the compression depth (CD), and for the thoracic impedance (TI). The objective was to evaluate their accuracy for compression detection and for CPR quality assessment. METHODS Compressions were manually annotated using the force and ECG in 38 out-of-hospital resuscitation episodes, comprising 869 min and 67,402 compressions. Compressions were detected using a negative peak detector for the CD. For the TI, an adaptive peak detector based on the amplitude and duration of TI fluctuations was used. Chest compression rate (CC-rate) and chest compression fraction (CCF) were calculated for the episodes and for every minute within each episode. CC-rate for rescuer feedback was calculated every 8 consecutive compressions. RESULTS The sensitivity and positive predictive value were 98.4% and 99.8% using CD, and 94.2% and 97.4% using TI. The mean CCF and CC-rate obtained from both detectors showed no significant differences with those obtained from the annotations (P>0.6). The Bland-Altman analysis showed acceptable 95% limits of agreement between the annotations and the detectors for the per-minute CCF, per-minute CC-rate, and CC-rate for feedback. For the detector based on TI, only 3.7% of CC-rate feedbacks had an error larger than 5%. CONCLUSION Automatic compression detectors based on the CD and TI signals are very accurate. In most cases, episode review could safely rely on these detectors without resorting to manual review. Automatic feedback on rate can be accurately done using the impedance channel.

Reliability and accuracy of the thoracic impedance signal for measuring cardiopulmonary resuscitation quality metrics

AIM To determine the accuracy and reliability of the thoracic impedance (TI) signal to assess cardiopulmonary resuscitation (CPR) quality metrics. METHODS A dataset of 63 out-of-hospital cardiac arrest episodes containing the compression depth (CD), capnography and TI signals was used. We developed a chest compression (CC) and ventilation detector based on the TI signal. TI shows fluctuations due to CCs and ventilations. A decision algorithm classified the local maxima as CCs or ventilations. Seven CPR quality metrics were computed: mean CC-rate, fraction of minutes with inadequate CC-rate, chest compression fraction, mean ventilation rate, fraction of minutes with hyperventilation, instantaneous CC-rate and instantaneous ventilation rate. The CD and capnography signals were accepted as the gold standard for CC and ventilation detection respectively. The accuracy of the detector was evaluated in terms of sensitivity and positive predictive value (PPV). Distributions for each metric computed from the TI and from the gold standard were calculated and tested for normality using one sample Kolmogorov-Smirnov test. For normal and not normal distributions, two sample t-test and Mann-Whitney U test respectively were applied to test for equal means and medians respectively. Bland-Altman plots were represented for each metric to analyze the level of agreement between values obtained from the TI and gold standard. RESULTS The CC/ventilation detector had a median sensitivity/PPV of 97.2%/97.7% for CCs and 92.2%/81.0% for ventilations respectively. Distributions for all the metrics showed equal means or medians, and agreements >95% between metrics and gold standard was achieved for most of the episodes in the test set, except for the instantaneous ventilation rate. CONCLUSION With our data, the TI can be reliably used to measure all the CPR quality metrics proposed in this study, except for the instantaneous ventilation rate.

Reliable extraction of the circulation component in the thoracic impedance measured by defibrillation pads.

AIM To analyze the feasibility of extracting the circulation component from the thoracic impedance acquired by defibrillation pads. The impedance circulation component (ICC) would permit detection of pulse-generating rhythms (PRs) during the analysis intervals of an automated external defibrillator when a non-shockable rhythm with QRS complexes is detected. METHODS A dataset of 399 segments, 165 associated with PR and 234 with pulseless electrical activity (PEA) rhythms, was extracted from out-of-hospital cardiac arrest episodes by applying a conservative criterion. Records consisted of the electrocardiogram and the thoracic impedance signals free of artifacts due to thoracic compressions and ventilations. The impedance was processed using an adaptive scheme based on a least mean square algorithm to extract the ICC. Waveform features of the ICC signal and its first derivative were used to discriminate PR from PEA rhythms. RESULTS The segments were split into development (83 PR and 117 PEA rhythms) and testing (82 PR and 117 PEA rhythms) subsets with a mean duration of 10.6s. Three waveform features, peak-to-peak amplitude, mean power, and mean area were defined for the ICC signal and its first derivative. The discriminative power in terms of area under the curve with the testing dataset was 0.968, 0.971, and 0.969, respectively, when applied to the ICC signal, and 0.974, 0.988 and 0.988, respectively, with its first derivative. CONCLUSION A reliable method to extract the ICC of the thoracic impedance is feasible. Waveform features of the ICC or its first derivative show a high discriminative power to differentiate PR from PEA rhythms (area under the curve higher than 0.96 for any feature).

A Reliable Method for Rhythm Analysis during Cardiopulmonary Resuscitation

Interruptions in cardiopulmonary resuscitation (CPR) compromise defibrillation success. However, CPR must be interrupted to analyze the rhythm because although current methods for rhythm analysis during CPR have high sensitivity for shockable rhythms, the specificity for nonshockable rhythms is still too low. This paper introduces a new approach to rhythm analysis during CPR that combines two strategies: a state-of-the-art CPR artifact suppression filter and a shock advice algorithm (SAA) designed to optimally classify the filtered signal. Emphasis is on designing an algorithm with high specificity. The SAA includes a detector for low electrical activity rhythms to increase the specificity, and a shock/no-shock decision algorithm based on a support vector machine classifier using slope and frequency features. For this study, 1185 shockable and 6482 nonshockable 9-s segments corrupted by CPR artifacts were obtained from 247 patients suffering out-of-hospital cardiac arrest. The segments were split into a training and a test set. For the test set, the sensitivity and specificity for rhythm analysis during CPR were 91.0% and 96.6%, respectively. This new approach shows an important increase in specificity without compromising the sensitivity when compared to previous studies.

Feasibility of automated rhythm assessment in chest compression pauses during cardiopulmonary resuscitation

AIM To demonstrate the feasibility of doing a reliable rhythm analysis in the chest compression pauses (e.g. pauses for two ventilations) during cardiopulmonary resuscitation (CPR). METHODS We extracted 110 shockable and 466 nonshockable segments from 235 out-of-hospital cardiac arrest episodes. Pauses in chest compressions were already annotated in the episodes. We classified pauses as ventilation or non-ventilation pause using the transthoracic impedance. A high-temporal resolution shock advice algorithm (SAA) that gives a shock/no-shock decision in 3s was launched once for every pause longer than 3s. The sensitivity and specificity of the SAA for the analyses during the pauses were computed. RESULTS We identified 4476 pauses, 3263 were ventilation pauses and 2183 had two ventilations. The median of the mean duration per segment of all pauses and of pauses with two ventilations were 6.1s (4.9-7.5s) and 5.1s (4.2-6.4s), respectively. A total of 91.8% of the pauses and 95.3% of the pauses with two ventilations were long enough to launch the SAA. The overall sensitivity and specificity were 95.8% (90% low one-sided CI, 94.3%) and 96.8% (CI, 96.2%), respectively. There were no significant differences between the sensitivities (P=0.84) and the specificities (P=0.18) for the ventilation and the non-ventilation pauses. CONCLUSION Chest compression pauses are frequent and of sufficient duration to launch a high-temporal resolution SAA. During these pauses rhythm analysis was reliable. Pre-shock pauses could be minimised by analysing the rhythm during ventilation pauses when CPR is delivered at 30:2 compression:ventilation ratio.

Can thoracic impedance monitor the depth of chest compressions during out-of-hospital cardiopulmonary resuscitation?

AIM To analyze the relationship between the depth of the chest compressions and the fluctuation caused in the thoracic impedance (TI) signal in out-of-hospital cardiac arrest (OHCA). The ultimate goal was to evaluate whether it is possible to identify compressions with inadequate depth using information of the TI waveform. METHODS 60 OHCA episodes were extracted, one per patient, containing both compression depth (CD) and TI signals. Every 5s the mean value of the maxima of the CD, Dmax, and three features characterizing the fluctuations caused by the compressions in the TI waveform (peak-to-peak amplitude, area and curve length) were computed. The linear relationship between Dmax and the TI features was tested using Pearson correlation coefficient (r) and univariate linear regression for the whole population, for each patient independently, and for series of compressions provided by a single rescuer. The power of the three TI features to classify each 5s-epoch as shallow/non-shallow was evaluated in terms of area under the curve, sensitivity and specificity. RESULTS The r was 0.34, 0.36 and 0.37 for peak-to-peak amplitude, area and curve length respectively when the whole population was analyzed. Within patients the median r was 0.40, 0.43 and 0.47, respectively. The analysis of the series of compressions yielded a median r of 0.81 between Dmax and the peak-to-peak amplitude, but it decreased to 0.47 when all the series were considered jointly. The classifier based on the TI features showed 90.0%/37.1% and 86.2%/43.5% sensitivity/specificity values, and an area under the curve of 0.75 and 0.71 for the training and test set respectively. CONCLUSION Low linearity between CD and TI was noted in OHCA episodes involving multiple rescuers. Our findings suggest that TI is unreliable as a predictor of Dmax and inaccurate in detecting shallow compressions.

A New Method for Feedback on the Quality of Chest Compressions during Cardiopulmonary Resuscitation

Quality of cardiopulmonary resuscitation (CPR) improves through the use of CPR feedback devices. Most feedback devices integrate the acceleration twice to estimate compression depth. However, they use additional sensors or processing techniques to compensate for large displacement drifts caused by integration. This study introduces an accelerometer-based method that avoids integration by using spectral techniques on short duration acceleration intervals. We used a manikin placed on a hard surface, a sternal triaxial accelerometer, and a photoelectric distance sensor (gold standard). Twenty volunteers provided 60 s of continuous compressions to test various rates (80–140 min−1), depths (3–5 cm), and accelerometer misalignment conditions. A total of 320 records with 35312 compressions were analysed. The global root-mean-square errors in rate and depth were below 1.5 min−1 and 2 mm for analysis intervals between 2 and 5 s. For 3 s analysis intervals the 95% levels of agreement between the method and the gold standard were within −1.64–1.67 min−1 and −1.69–1.72 mm, respectively. Accurate feedback on chest compression rate and depth is feasible applying spectral techniques to the acceleration. The method avoids additional techniques to compensate for the integration displacement drift, improving accuracy, and simplifying current accelerometer-based devices.