Prediction of countershock success using single features from multiple ventricular fibrillation frequency bands and feature combinations using neural networks.
ABSTRACT Targeted defibrillation therapy is needed to optimise survival chances of ventricular fibrillation (VF) patients, but at present VF analysis strategies to optimise defibrillation timing have insufficient predictive power. From 197 patients with in-hospital and out-of-hospital cardiac arrest, 770 electrocardiogram (ECG) recordings of countershock attempts were analysed. Preshock VF ECG features in the time and frequency domain were tested retrospectively for outcome prediction. Using band pass filters, the ECG spectrum was split into various frequency bands of 2-26 Hz bandwidth in the range of 0-26 Hz. Neural networks were used for single feature combinations to optimise prediction of countershock success. Areas under curves (AUC) of receiver operating characteristics (ROC) were used to estimate prediction power of single and combined features. The highest ROC AUC of 0.863 was reached by the median slope in the interval 10-22 Hz resulting in a sensitivity of 95% and a specificity of 50%. The best specificity of 55% at the 95% sensitivity level was reached by power spectrum analysis (PSA) in the 6-26 Hz interval. Neural networks combining single predictive features were unable to increase outcome prediction. Using frequency band segmentation of human VF ECG, several single predictive features with high ROC AUC>0.840 were identified. Combining these single predictive features using neural networks did not further improve outcome prediction in human VF data. This may indicate that various simple VF features, such as median slope already reach the maximum prediction power extractable from VF ECG.
- SourceAvailable from: Trygve Eftestøl[show abstract] [hide abstract]
ABSTRACT: 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.BioMed research international. 01/2014; 2014:386010.
- [show abstract] [hide abstract]
ABSTRACT: The capability of Amplitude Spectrum Area (AMSA) to predict the success of defibrillation (DF) was retrospectively evaluated in a large database of out-of-hospital cardiac arrests. Electrocardiographic data, including 1260 DFs, were obtained from 609 cardiac arrest patients due to ventricular fibrillation. AMSA sensitivity, specificity, accuracy, and positive and negative predictive values (PPV, NPV) for predicting DF success were calculated, together with receiver operating characteristic (ROC) curves. Successful DF was defined as the presence of spontaneous rhythm ≥ 40 bpm starting within 60sec from the DF. In 303 patients with chest compression (CC) depth data collected with an accelerometer, changes in AMSA were analyzed in relationship to CC depth. AMSA was significantly higher prior to a successful DF than prior to an unsuccessful DF (15.6±0.6 vs. 7.97±0.2 mV-Hz, p<0.0001). Intersection of sensitivity, specificity and accuracy curves identified a threshold AMSA of 10 mV-Hz to predict DF success with a balanced sensitivity, specificity and accuracy of almost 80%. Higher AMSA thresholds were associated with further increases in accuracy, specificity and PPV. AMSA of 17 mV-Hz predicted DF success in two third of instances (PPV of 67%). Low AMSA, instead, predicted unsuccessful DFs with high sensitivity and NPV >97%. Area under the ROC curve was 0.84. CC depth affected AMSA value. When depth was <1.75 in, AMSA decreased for consecutive DFs, while it increased when the depth was >1.75 in (p<0.05). AMSA could be a useful tool to guide CPR interventions and predict the optimal timing of DF.Resuscitation 09/2013; · 4.10 Impact Factor
- Resuscitation 10/2013; · 4.10 Impact Factor