Beruflich Dokumente
Kultur Dokumente
Background—In 156 patients with out-of-hospital cardiac arrest of cardiac cause, we analyzed the ability of 4 spectral
features of ventricular fibrillation before a total of 868 shocks to discriminate or not between segments that correspond
to return of spontaneous circulation (ROSC).
Methods and Results—Centroid frequency, peak power frequency, spectral flatness, and energy were studied. A second
decorrelated feature set was generated with the coefficients of the principal component analysis transformation of the
original feature set. Each feature set was split into training and testing sets for improved reliability in the evaluation of
nonparametric classifiers for each possible feature combination. The combination of centroid frequency and peak power
frequency achieved a mean⫾SD sensitivity of 92⫾2% and specificity of 27⫾2% in testing. The highest performing
classifier corresponded to the combination of the 2 dominant decorrelated spectral features with sensitivity and
specificity equal to 92⫾2% and 42⫾1% in testing or a positive predictive value of 0.15 and a negative predictive value
of 0.98. Using the highest performing classifier, 328 of 781 shocks not leading to ROSC would have been avoided,
whereas 7 of 87 shocks leading to ROSC would not have been administered.
Conclusions—The ECG contained information predictive of shock therapy. This could reduce the delivery of unsuccessful
shocks and thereby the duration of unnecessary “hands-off” intervals during cardiopulmonary resuscitation. The low
specificity and positive predictive value indicate that other features should be added to improve performance.
(Circulation. 2000;102:1523-1529.)
Key Words: cardiopulmonary resuscitation 䡲 fibrillation 䡲 defibrillation 䡲 Fourier analysis
Received February 7, 2000; revision received April 26, 2000; accepted May 2, 2000.
From Høgskolen i Stavanger (T.E., S.O.A., J.H.H.), Department of Electrical and Computer Engineering, Stavanger, Norway; Ulleval University
Hospital (K.S.), Institute for Experimental Medical Research and Norwegian Air Ambulance, Oslo, Norway; and Ulleval University Hospital (K.S.,
P.A.S.), Department of Anesthesiology, Oslo, Norway.
Correspondence to Trygve Eftestøl, Høgskolen i Stavanger, Department of Electrical and Computer Engineering, PO Box 2557, Ullandhaug, 4091
Stavanger, Norway. E-mail trygve-e@ux.his.no
© 2000 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
1523
1524 Circulation September 26, 2000
冏冘 冏
TABLE 1. Class Division Scheme L⫺1
2
⫺j2 fn
Class Subclass Conversion to n Prior ACLS, % P̂共f兲⫽ x共n兲e
n⫽0
ROSC 1 Pulse rhythm 87 74
where f is the frequency and x(n) denotes sample n in an ECG
No-ROSC 2 Pulseless rhythm 337 86
segment of length L.
No-ROSC 3 Isoelectric 98 66 We attempted to discriminate between preshock ECG segments
No-ROSC 4 VF 35 77 that correspond to ROSC and No ROSC outcome by computing the
following features from the ECG segment PSD estimates.
No-ROSC 5 Non-reset shocks 311 87 The CF, or median frequency,4,5,15,17 is given by
冕
Of the 156 patients, 110 received shock treatment. The 87 shocks in the
fu
ROSC class were administered to 46 of the patients. Of these shocks, 35
converted to sustained pulse rhythm, whereas 52 converted to pulse rhythm of
fP̂共f兲df
fl
limited duration (⬎20 seconds). Nineteen ROSCs were achieved after the first
冕
CF⫽
shock (4 with prior ACLS). fu
P̂共f兲df
fl
data set (“testing set”) instead of both having been deter-
mined from the same data set. where f1ⱕfⱕfu; f1 and fu are the lower- and higher-frequency band
limits, respectively. By varying these limits, we could study the
We have therefore attempted to predict defibrillation out- effect of extracting features from different frequency bands.
come in human cardiac arrest by combining features of PPF is given by
spectral characterization, splitting the data into training and
testing sets, and using classifier generalization techniques in PPF⫽arg max关P̂共f兲兴
f
an attempt to increase the degree of expected reliability. One
of the combinations studied was that reported by Brown et The spectral flatness measure (SFM)19 of the VF is given by:
al.12
冕 e fl
fu
ln P̂共 f兲 df
e冕
Methods SFM⫽2
fu
P̂共 f兲 df
In the observational prospective study from Oslo,11 data were fl
collected from the medical control module of the defibrillator
(Heartstart 3000; Laerdal Medical) and the regular Utstein registra- SFM attains a value between 0 (peaky) and 1 (flat).
tion.13 All patients with out-of-hospital cardiac arrest of cardiac Various time domain measurements of signal amplitude charac-
origin13 that occurred between February 19, 1996, and February 18, teristics of VF have been investigated.6,12,16,20 –22 In the present
1998, were included if the advanced cardiac life support (ACLS) study, we investigated an alternative frequency band–limited energy
attempt was documented on the medical control module. Approval measurement (ENRG):
冕
for the study was obtained through the Regional Committee for fu
Research Ethics, Health Region III (Norway), and the Norwegian ENRG⫽ P̂共f兲df
Data Inspectorate.
fl
冘
Feature Extraction K
TABLE 2. Tested Combinations of Candidate Features specificity (probability of negative prediction of No ROSC outcome)
given by
Combination Spectral PCA
Psns共 i 兲⫽P共Ri 兩 i 兲
v v
1 [SFM] [PCA1] and
2
3
[ENRG]
[CF]
[PCA2]
[PCA3]
Pspc共 i 兲⫽
1
1⫺P共 i 兲 冘
j,k⫽i
P共 k 兲P共Rj 兩 k 兲
4 [PPF] [PCA4]
respectively. P(Rj兩i) expresses the proportion of true class i with
5 [SFM ENRG] [PCA1 PCA2]
the corresponding decision being ˆ j.
6 [SFM CF] [PCA1 PCA3] The decision regions were calculated iteratively with minimiza-
7 [SFM PPF] [PCA1 PCA4] tion of the object function
8 [ENRG CF] [PCA2 PCA3] J⫽关Psnsd共 i 兲⫺Psns共 i 兲兴 2
9 [ENRG PPF] [PCA2 PCA4] so that the classifier would meet the desired performance criterion
10 [CF PPF] [PCA3 PCA4] given by Psnsd(i). This is done by multiplying the costs, C(i, ˆ j),
11 [SFM ENRG CF] [PCA1 PCA2 PCA3] j⫽i, by factor ␣. By setting i⫽1 (1 corresponding to ROSC), this
allowed specification of a sensitivity for the recognition of ROSC
12 [SFM ENRG PPF] [PCA1 PCA2 PCA4] outcome.
13 [SFM CF PPF] [PCA1 PCA3 PCA4] The underlying statistics can be estimated with classification
14 [ENRG CF PPF] [PCA2 PCA3 PCA4] theory.14 Multidimensional histograms were applied in which the
feature space is divided into bins of equal volume, in which the PDF
15 [SFM ENRG CF PPF] [PCA1 PCA2 PCA3 PCA4] estimates are computed. Each feature set is normalized by dividing
Shown are vectors for SFM, ENRG, CF, and PPF and their respective by the respective feature axis into nb equal-sized intervals in the
decorrelated features by PCA transformation. range from the minimum to the maximum feature value. The PDF
estimates in each histogram bin are then distributed by applying an
elliptic gaussian kernel function, resulting in a smoother continuous
This corresponds to minimization of the expectation of the classifier estimate.14
risk.14 P(j兩v), j⫽1, . . . , K denotes the a posteriori probability Histogram bin resolution and kernel width are the 2 key parame-
function for class j, which is derived according to Bayes’ rule: ters of the classifier. A small number of large bins provide low
histogram resolution, whereas a large number of small bins provide
P共 j 兲p共v兩 j 兲 high resolution. Each feature axis of the PDF is divided into nb
P共 j 兩v兲⫽ , j⫽1, . . . , K
冘
K intervals. Thus, if the feature dimension is D for a specific feature
P共 i 兲p共v兩 i 兲 combination, the feature space is divided into nbD bins of equal
i⫽1 volume. Smoothness is governed by the width of the kernel function.
A narrow kernel function provides a high-resolution estimate with
P(i) and p(v兩i) denote the a priori probability and the class-specific high variance, whereas a wide kernel function provides a smoother
probability density function (PDF) for class i, respectively. low-resolution estimate with low variance.
The classifier performance characteristics are expressed by the The concept of generality is important in the design of classifiers.
sensitivity (probability of positive prediction of ROSC outcome) and The decision regions are calculated with a training set of feature
vectors that represent the experience on which the classifier will base Three feature sets were extracted with frequency ranges (fl⫺fu Hz) of
future decisions. Testing is done on an independent set. In a 0 to 50, 0 to 25, and 0 to 12.5 Hz. The spectral features produced in
well-designed classifier, the testing performance should approach the each of these experiments were vSFM, vENRG, vCF, and vPPF. The PCA
training performance. Both the histogram bin resolution and the transformation of these features gave the corresponding decorrelated
kernel width applied in the estimation affect generality. feature set of vPCA1, vPCA2, vPCA3, and vPCA4. The ECG immediately
Training and testing were conducted with a cross-validation before defibrillation was analyzed, and the measurements were
technique.14 In each of S consecutive experiments, an (S⫺1)/S grouped according to the postshock rhythm for classifier design
portion of the entire data set is used to train classifier number i. The (Table 1).
remaining 1/S portion is kept out for testing. i is varied from 1 to S, Classifiers were designed and tested with the use of all possible
thus producing S classifier performance results. combinations of spectral features and decorrelated features (Table 2).
The statistical functions were estimated with multidimensional
Experimental Setup histograms. Resolutions were adjusted according to setting nb equal
to 4, 8, 16, 32, 64, and 128 bins. For each of these resolutions, the
The ECG was sampled at 100 Hz with 8-bit resolution, and PSD was smoothness was varied by setting the variance of the gaussian kernel
estimated from segment lengths L⫽400 zero padded to 512 samples. function, kw, equal to 0, 1, 5, 10, 15, and 20.
TABLE 6. Distribution of the Single Feature With Highest myocardial dysfunction.7 Moreover, because the spectral
Prediction Performance for Shocks Administered Before Any characteristics of the VF have been reported to reflect
ACLS vs After ACLS myocardial perfusion,4 – 6 the defibrillator also might guide
No ROSC ROSC P the CPR attempt, because the myocardial perfusion depends
on compression force, rate, and duration.24 –27
⬍0.0001*
On the other hand, 7 shocks that resulted in a pulse-giving
ACLS ⫺2438 (⫺5932, 2369) 6068 (956, 10 788) ⬍0.0001
rhythm would not have been administered. These shocks
No ACLS 708 (⫺4588, 5053) 4049 (⫺605, 9948) ⬍0.01 presumably would have been administered later if CPR
P ⬍0.05 ⬎0.05 ⬍0.001* changed the characteristics of the VF. The effects of this
P value for differences between subgroup features along horizontal, vertical, could not be evaluated. The comparison of ACLS with No
and diagonal* lines. ACLS features illustrates this aspect of use of the features for
online monitoring of the CPR efficiency. The use of the
fiers gave the best results for all frequency ranges. The features as monitoring parameters for performance feedback
combination of decorrelated features improved the perfor- during CPR is an interesting idea that is closely related to the
mance significantly when the 2 midfrequency-range principal prediction problem. Retrospectively, we studied the influence
features were combined. The inclusion of ⬎2 decorrelated of ACLS on a single feature and demonstrated changes in
features did not further improve the performance. values according to treatment. The present study demon-
Whether ACLS caused changes in the PCA1 feature is strates how a general classifier can be designed through
summarized in Table 6. The No ACLS/No ROSC subgroup cross-validation, which allows training and testing on inde-
may be considered the starting point, where the initial shocks pendent data sets in combination with different resolutions
are futile, and is further divided into the following subgroups: and kernel widths in the estimation of the statistics that
The ACLS/No ROSC subgroup, where treatment has been describe the features. This method gives an indication of how
futile and the myocardial condition probably has deteriorated well the classifier will perform when challenged with new
as reflected by a significant decrease in the feature values data in the future.
The ACLS/ROSC subgroup, where treatment probably has In a similar study of 128 shocks in 55 patients with only 9
caused an improvement in myocardial condition, which is successful shocks (defined as a conversion of VF to a
reflected by a significant increase in the feature values supraventricular rhythm with a palpable pulse or blood
comparable to that corresponding to the No ACLS/ROSC pressure of any duration within 2 minutes of the shock
subgroup. without ongoing CPR), Brown et al12 extracted 4 parameters
from the recorded ECG. The combination of CF and PPF
Discussion gave the best predictive potential (sensitivity 100%, specific-
In this study of 868 shocks in 156 patients, it was possible to ity 47.1%).12 The same combination of features gave a poorer
predict in part whether the shock resulted in ROSC or No predictive potential in the present study (sensitivity 92⫾2%,
ROSC by analyzing 4 spectral features of the preshock ECG specificity 27⫾2%). We believe that the results of our
with improved results by combining, decorrelating, and generalized classifier are more realistic due to the larger
reducing the features. database and the use of independent testing and generaliza-
We further demonstrated how classification methodology tion that were not done by Brown et al.12 Those authors
allows the combination of features with an increase in generated the sensitivity and specificity with the same data
classifier performance compared with individual classifica- from which the threshold values were computed, with no
tion of features. We also showed how decorrelation by PCA independent evaluation.
allows dimensional reduction in the feature set with no Noc et al6 reported in pigs that maximum and mean VF
decrease in performance compared with a combination of the amplitude and dominant VF frequency were all acceptable
complete feature set. shock outcome predictors. They derived the threshold values
The rate of ROSC after individual shocks in patients is from 1 group and tested these in a separate validation group
reported to be low8,9,11,12; the rate was 10% in a recent study but had different results in the 2 groups, indicating that the
from Oslo.11 Most shocks are thus individually futile. Based results might not be reliable.6 Our results indicate that Brown
on the present results, 42% of the unsuccessful shocks (328 of et al12 would have experienced the same if their threshold
781) could have been avoided, and a period of chest com- values had been tested on new data.
pressions, ventilations, and vasoactive drugs could have been Our method includes independent testing and generaliza-
administered before a new defibrillation attempt was made. tion to avoid these problems. To ensure reliability, the data
Studies in animals have shown that this may be favorable,23 were split in 2. Training performance for ROSC and No
and a recent study in humans indicated that this might ROSC prediction was computed from half of the data,
improve the outcome.3 It would minimize the detriment of whereas the other half was used to compute the correspond-
“hands-off” intervals, where the vital organs are without ing test performance.
perfusion, which reduces the possibility of ROSC, recovery There are some limitations in the present study. First, the
with intact neurological status, or both. The number of shocks number of ROSC observations is low. Second, in the cross-
should also be kept to a minimum, because repetitive shocks validation processing of the data, the test performances were
and total electric power are injurious to the already ischemic considered in the design of the classifiers to choose the
myocardium and increase the severity of postresuscitation generalizing parameters. Ideally, a final evaluation should
Eftestøl et al Outcome of Defibrillation in Out-of-Hospital Cardiac Arrest 1529
have been performed on yet another data set that did not 10. Gliner BE, White RD. Electrocardiographic evaluation of defibrillation
influence the design process. Third, we used only 1 type of shocks delivered to out-of- hospital sudden cardiac arrest patients. Resus-
citation. 1999;41:133–144.
classifier: the histogram method. To obtain even more reli- 11. Sunde K, Eftestøl T, Aschenberg C, et al. Quality assessment of defibril-
able results, the experiments should be repeated with other lation and ALS using data from the medical control module of the
types of classifiers. defibrillator. Resuscitation. 1999;41:237–247.
12. Brown CG, Dzwonczyk R. Signal analysis of the human electrocardio-
Spectral characterization of VF can be of clinical impor- gram during ventricular fibrillation: frequency and amplitude parameters
tance if it can be incorporated into the software of defibril- as predictors of successful shock. Ann Emerg Med. 1996;17:436 – 437.
lators. We have demonstrated a method to develop an 13. Chamberlain D, Cummins RO, Abramson NS, et al. Recommended
outcome predictor for defibrillation attempts in out-of- guidelines for uniform reporting of data from out-of-hospital cardiac
arrest: the “Utstein style.” Resuscitation. 1991;22:1–26.
hospital cardiac arrest patients, although the sensitivity of 14. Schurmann J. Pattern Classification: A Unified View of Statistical and
92⫾2% and specificity of 42⫾1% are not satisfactory for Neural Approaches. New York, NY: John Wiley & Sons; 1996.
clinical use. Therefore, other features should also be investi- 15. Brown CG, Dzwonczyk R, Werman HA, et al. Estimating the duration of
ventricular fibrillation. Ann Emerg Med. 1989;18:1181–1185.
gated to add discriminative power to the feature set. 16. Stewart AJ, Allen JD, Adgey AAJ. Frequency analysis of ventricular
fibrillation and resuscitation success. Q J Med. 1992;85:761–769.
Acknowledgments 17. Strohmenger HU, Lindnder KH, Lurie KG, et al. Frequency of ventricular
This work was supported in part by the Research Council of Norway, fibrillation as a predictor of defibrillation success during cardiac surgery.
the Norwegian Air Ambulance, the Laerdal Foundation for Acute Anesth Analg. 1994;79:434 – 438.
Medicine, and Anders Jahre’s Foundation. 18. Kay SM. Modern Spectral Estimation: Theory & Application. Englewood
Cliffs, NJ: Prentice-Hall; 1988.
19. Jayant NS, Noll P. Digital Coding of Waveforms: Principles and Appli-
References cations to Speech and Video. Englewood Cliffs, NJ: Prentice-Hall; 1984.
1. Larsen MP, Eisenberg MS, Cummins RO, et al. Predicting survival from 20. Weaver WD, Cobb LA, Dennis D, et al. Amplitude of ventricular fibril-
out-of-hospital cardiac arrest: a graphic model. Ann Emerg Med. 1993; lation waveform and outcome after cardiac arrest. Ann Intern Med.
22:1652–1658. 1985;102:53–55.
2. Stiell IG, Wells GA, DeMaio VJ, et al. Modifiable factors associated with 21. Callaham M, Braun O, Valentine W, et al. Prehospital cardiac arrest
improved cardiac arrest survival in a multicenter basic life support/ treated by urban first-responders: profile of patient response and pre-
defibrillation system: OPALS Study Phase I Results. Ann Emerg Med. diction of outcome by ventricular fibrillation waveform. Ann Emerg Med.
1999;33:44 –50. 1993;22:1664 –1677.
3. Cobb LA, Fahrenbruch CE, et al. Influence of cardiopulmonary resusci- 22. Kuelz KW, Hsia P, Wise RM, et al. Integration of absolute ventricular
tation prior to defibrillation in patients with out-of-hospital ventricular fibrillation voltage correlates with successful defibrillation. IEEE Trans
fibrillation. JAMA. 1999;28:1182–1188. BME. 1994;41:782–791.
4. Brown CG, Griffith RF, Ligten PV, et al. Median frequency: a new 23. Niemann JT, Cairns CB, Sharma J, et al. Treatment of prolonged ven-
parameter for predicting defibrillation success rate. Ann Emerg Med. tricular fibrillation: immediate countershock versus high-dose epineph-
1991;20:787–789. rine and CPR preceding countershock. Circulation. 1992;85:281–287.
5. Strohmenger HU, Lindnder KH, Lindner IM, et al. Spectral analysis of 24. Maier GW, Tyson GS Jr, Olsen CO, et al. The physiology of external
ventricular fibrillation and closed-chest cardiopulmonary resuscitation. cardiac massage: high-impulse cardiopulmonary resuscitation. Circu-
Resuscitation. 1996;33:155–161. lation. 1984;70:86 –101.
6. Noc M, Weill MH, Tang W, et al. Electrocardiographic prediction of the 25. Halperin HR, Tsitlik JE, Guerci AD, et al. Determinants of blood flow to
success of cardiac resuscitation. Crit Care Med. 1999;27:708 –714. vital organs during cardiopulmonary resuscitation in dogs. Circulation.
7. Xie J, Weil MH, Sun S, et al. High-energy defibrillation increases the 1986;73:539 –550.
severity of postresuscitation myocardial dysfunction. Circulation. 1997; 26. Ornato JP, Levine RL, Young DS, et al. The effect of applied chest
96:683– 688. compression force on systemic arterial pressure and end-tidal carbon
8. Hargarten KM, Stueven HA, Waite EM, et al. Prehospital experience with dioxide concentration during CPR in human beings. Ann Emerg Med.
defibrillation of coarse ventricular fibrillation: a ten-year review. Ann 1989;18:732–737.
Emerg Med. 1990;19:157–162. 27. Feneley MP, Maier GW, Kern KB, et al. Influence of compression rate on
9. Behr JC, Harley LL, York DK, et al. truncated exponential versus damped initial success of resuscitation and 24 hour survival after prolonged
sinusoidal waveform shocks for transthoracic defibrillation. Am J manual cardiopulmonary resuscitation in dogs. Circulation.
Cardiol. 1996;78:1242–1245. 1988;77:240 –250.