Abstract 202: Predicting Successful Defibrillation with a New Algorithm Based on the Magnitude and Dominant Frequency of Ventricular Fibrillation Waveform
After prolonged ventricular fibrillation (VF), defibrillation after every 2 minutes of CPR, as currently recommended, not only often fails to return spontaneous circulation but also causes significant myocardial injury. This study has developed a new algorithm to predict the optimal time of defibrillation. We hypothesized that the product of the total absolute magnitudes and dominant frequency (PMD) will provide an equal or improved prediction of amplitude spectrum area (AMSA) in a pig model of CPR. Forty-eight domestic pigs weighing between 35 to 48 kg were utilized. The durations of untreated ventricular fibrillation (VF) were from 4 to 15 minutes. The electrical shock was delivered after 2 to 5 minutes of chest compression. Only the results of the first electrical shock were reported. The 5-sec strip of VF waveform prior to each electrical shock was captured to calculate PMD and AMSA. The formula calculating PMD is PMD= (|A1|+|A2|+¼+|Ai|+¼+|An|) * DF/K, where |Ai| is the absolute amplitude of the ith time-sampled data. DF is the dominant frequency. K is a constant factor. For PMD calculation, the frequency range from 5 Hz to 25 Hz was selected to filter the signal, for AMSA calculation, the frequently-used frequency range is from 4 Hz to 48Hz. K=100 was used to calculate PMD. The perfusing rhythm was restored in 20 pigs after the electrical shock, PMD was 22.95±5.01 mV·Hz in resuscitated pigs and was 10.21±3.55 mV·Hz when electrical shock failed to restore circulation (p<.001). AMSA values were 30.97±5.35 mV·Hz and 18.77±4.57 mV·Hz in resuscitated and non-resuscitated pigs, respectively (p<.001). The differences in both of PMD and AMSA between resuscitated and non-resuscitated pigs were highly significant (p<.001). For PMD, the calculated threshold was 16 mV·Hz when the sensitivity and specificity were 100% and 89%, respectively. For AMSA, the calculated threshold was 23 mV·Hz when the sensitivity and specificity were 100% and 86%, respectively. We concluded that the product of the total absolute magnitudes and dominant frequency (PMD) predicts the successful defibrillation during CPR as effectively as amplitude spectrum area (AMSA).
- © 2011 by American Heart Association, Inc.