1
|
Ivanović MD, Hannink J, Ring M, Baronio F, Vukčević V, Hadžievski L, Eskofier B. Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design. Artif Intell Med 2020; 110:101963. [PMID: 33250144 DOI: 10.1016/j.artmed.2020.101963] [Citation(s) in RCA: 9] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/11/2019] [Revised: 08/23/2020] [Accepted: 09/22/2020] [Indexed: 10/23/2022]
Abstract
OBJECTIVE Optimizing timing of defibrillation by evaluating the likelihood of a successful outcome could significantly enhance resuscitation. Previous studies employed conventional machine learning approaches and hand-crafted features to address this issue, but none have achieved superior performance to be widely accepted. This study proposes a novel approach in which predictive features are automatically learned. METHODS A raw 4s VF episode immediately prior to first defibrillation shock was feed to a 3-stage CNN feature extractor. Each stage was composed of 4 components: convolution, rectified linear unit activation, dropout and max-pooling. At the end of feature extractor, the feature map was flattened and connected to a fully connected multi-layer perceptron for classification. For model evaluation, a 10 fold cross-validation was employed. To balance classes, SMOTE oversampling method has been applied to minority class. RESULTS The obtained results show that the proposed model is highly accurate in predicting defibrillation outcome (Acc = 93.6 %). Since recommendations on classifiers suggest at least 50 % specificity and 95 % sensitivity as safe and useful predictors for defibrillation decision, the reported sensitivity of 98.8 % and specificity of 88.2 %, with the analysis speed of 3 ms/input signal, indicate that the proposed model possesses a good prospective to be implemented in automated external defibrillators. CONCLUSIONS The learned features demonstrate superiority over hand-crafted ones when performed on the same dataset. This approach benefits from being fully automatic by fusing feature extraction, selection and classification into a single learning model. It provides a superior strategy that can be used as a tool to guide treatment of OHCA patients in bringing optimal decision of precedence treatment. Furthermore, for encouraging replicability, the dataset has been made publicly available to the research community.
Collapse
Affiliation(s)
- Marija D Ivanović
- Vinca Institute of Nuclear Scientists, University of Belgrade, Belgrade, Serbia.
| | - Julius Hannink
- Machine Learning and Data Analytics Lab, Department of Computer Science, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Germany
| | - Matthias Ring
- Machine Learning and Data Analytics Lab, Department of Computer Science, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Germany
| | - Fabio Baronio
- CNR and Department of Information Engineering, University of Brescia, Brescia, Italy
| | - Vladan Vukčević
- School of Medicine, University of Belgrade, Belgrade, Serbia
| | - Ljupco Hadžievski
- Vinca Institute of Nuclear Scientists, University of Belgrade, Belgrade, Serbia; Diasens, Belgrade, Serbia
| | - Bjoern Eskofier
- Machine Learning and Data Analytics Lab, Department of Computer Science, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Germany
| |
Collapse
|
2
|
Chen WT, Tsai MS, Tsai SH, Jiang YCF, Yang TJ, Huang CH, Chang WT, Chen WJ. Frequency Variation of Ventricular Fibrillation May Help Predict Successful Defibrillation in a Rat Model of Cardiac Arrest. J Acute Med 2019; 9:49-58. [PMID: 32995231 PMCID: PMC7440373 DOI: 10.6705/j.jacme.201906_9(2).0002] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/03/2018] [Revised: 07/27/2018] [Accepted: 08/16/2018] [Indexed: 06/11/2023]
Abstract
BACKGROUND To evaluate whether the frequency variation of ventricular fibrillation (VF) helps to predict successful defibrillation in a rat model of cardiac arrest. METHODS VF was induced in rats followed by cardiopulmonary resuscitation and then defibrillation. The electrocardiographic signals of 30 rats with first-shock success were obtained from our previous animal experiments, and 300 rats without first-shock success were selected as control. The VF waveform immediately before the first defibrillation was analyzed. RESULTS Eighty-eight percentages of the frequency variations of an electrocardiogram (ECG) record falling in the range -9.5-9.5 Hz was selected with sensitivity of 0.8, specificity of 0.583, and area under curve (AUC) of 0.708. Compared with amplitude spectrum area (AMSA) (sensitivity = 0.767, specificity= 0.547, and AUC = 0.678), combining frequency variation and AMSA significantly increases the predictability with sensitivity of 0.933, specificity of 0.493, and AUC of 0.732 (p = 0.005). CONCLUSIONS The frequency variation of VF may serve a useful parameter to predict defibrillation success.
Collapse
Affiliation(s)
- Wei-Ting Chen
- National Taiwan University Medical College and HospitalDepartment of Emergency MedicineTaipeiTaiwan
| | - Min-Shan Tsai
- National Taiwan University Hospital Hsin-Chu BranchDepartment of Emergency MedicineHsinchuTaiwan
| | - Shang-Ho Tsai
- National Chiao Tung UniversityDepartment of Electrical EngineeringHsinchuTaiwan
| | - Yu-Chen Fang Jiang
- National Chiao Tung UniversityDepartment of Electrical EngineeringHsinchuTaiwan
| | - Teck-Jin Yang
- Sijhih Cathay General HospitalDepartment of Emergency MedicineTaipeiTaiwan
| | - Chien-Hua Huang
- National Taiwan University Medical College and HospitalDepartment of Emergency MedicineTaipeiTaiwan
| | - Wei-Tien Chang
- National Taiwan University Medical College and HospitalDepartment of Emergency MedicineTaipeiTaiwan
| | - Wen-Jone Chen
- National Chiao Tung UniversityDepartment of Electrical EngineeringHsinchuTaiwan
| |
Collapse
|
3
|
Coult J, Blackwood J, Sherman L, Rea TD, Kudenchuk PJ, Kwok H. Ventricular Fibrillation Waveform Analysis During Chest Compressions to Predict Survival From Cardiac Arrest. Circ Arrhythm Electrophysiol 2019; 12:e006924. [PMID: 30626208 PMCID: PMC6532650 DOI: 10.1161/circep.118.006924] [Citation(s) in RCA: 21] [Impact Index Per Article: 4.2] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 01/08/2023]
Abstract
BACKGROUND Quantitative measures of the ventricular fibrillation (VF) ECG waveform can assess myocardial physiology and predict cardiac arrest outcomes, making these measures a candidate to help guide resuscitation. Chest compressions are typically paused for waveform measure calculation because compressions cause ECG artifact. However, such pauses contradict resuscitation guideline recommendations to minimize cardiopulmonary resuscitation interruptions. We evaluated a comprehensive group of VF measures with and without ongoing compressions to determine their performance under both conditions for predicting functionally-intact survival, the study's primary outcome. METHODS Five-second VF ECG segments were collected with and without chest compressions before 2755 defibrillation shocks from 1151 out-of-hospital cardiac arrest patients. Twenty-four individual measures and 3 combination measures were implemented. Measures were optimized to predict functionally-intact survival (Cerebral Performance Category score ≤2) using 460 training cases, and their performance evaluated using 691 independent test cases. RESULTS Measures predicted functionally-intact survival on test data with an area under the receiver operating characteristic curve ranging from 0.56 to 0.75 (median, 0.73) without chest compressions and from 0.53 to 0.75 (median, 0.69) with compressions ( P<0.001 for difference). Of all measures evaluated, the support vector machine model ranked highest both without chest compressions (area under the receiver operating characteristic curve, 0.75; 95% CI, 0.73-0.78) and with compressions (area under the receiver operating characteristic curve, 0.75; 95% CI, 0.72-0.78; P=0.75 for difference). CONCLUSIONS VF waveform measures predict functionally-intact survival when calculated during chest compressions, but prognostic performance is generally reduced compared with compression-free analysis. However, support vector machine models exhibited similar performance with and without compressions while also achieving the highest area under the receiver operating characteristic curve. Such machine learning models may, therefore, offer means to guide resuscitation during uninterrupted cardiopulmonary resuscitation.
Collapse
Affiliation(s)
- Jason Coult
- Department of Bioengineering, University of Washington,
Seattle, WA
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
| | - Jennifer Blackwood
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
- King County Emergency Medical Services, Seattle King County
Department of Public Health, Seattle, WA
| | - Lawrence Sherman
- Department of Bioengineering, University of Washington,
Seattle, WA
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
- Department of Medicine, University of Washington School of
Medicine, Seattle, WA
| | - Thomas D. Rea
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
- King County Emergency Medical Services, Seattle King County
Department of Public Health, Seattle, WA
- Department of Medicine, University of Washington School of
Medicine, Seattle, WA
| | - Peter J. Kudenchuk
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
- King County Emergency Medical Services, Seattle King County
Department of Public Health, Seattle, WA
- Division of Cardiology, University of Washington School of
Medicine, Seattle, WA
| | - Heemun Kwok
- Center for Progress in Resuscitation, University of
Washington, Seattle, WA
- Department of Emergency Medicine, University of Washington
School of Medicine, Seattle, WA
| |
Collapse
|
4
|
Ivanović MD, Ring M, Baronio F, Calza S, Vukčević V, Hadžievski L, Maluckov A, Eskofier B. ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients. Biomed Phys Eng Express 2018. [DOI: 10.1088/2057-1976/aaebec] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/12/2022]
|
5
|
Chicote B, Irusta U, Aramendi E, Alcaraz R, Rieta JJ, Isasi I, Alonso D, Baqueriza MDM, Ibarguren K. Fuzzy and Sample Entropies as Predictors of Patient Survival Using Short Ventricular Fibrillation Recordings during out of Hospital Cardiac Arrest. ENTROPY (BASEL, SWITZERLAND) 2018; 20:E591. [PMID: 33265680 PMCID: PMC7513119 DOI: 10.3390/e20080591] [Citation(s) in RCA: 11] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 07/06/2018] [Revised: 08/04/2018] [Accepted: 08/07/2018] [Indexed: 12/23/2022]
Abstract
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.
Collapse
Affiliation(s)
- Beatriz Chicote
- Department of Communications Engineering, University of the Basque Country (UPV/EHU), 48013 Bilbao, Spain
| | - Unai Irusta
- Department of Communications Engineering, University of the Basque Country (UPV/EHU), 48013 Bilbao, Spain
| | - Elisabete Aramendi
- Department of Communications Engineering, University of the Basque Country (UPV/EHU), 48013 Bilbao, Spain
| | - Raúl Alcaraz
- Research Group in Electronic, Biomedical and Telecommunication Engineering, University of Castilla-La Mancha (UCLM), 16071 Cuenca, Spain
| | - José Joaquín Rieta
- BioMIT.org, Electronic Engineering Department, Universitat Politécnica de Valencia (UPV), 46022 Valencia, Spain
| | - Iraia Isasi
- Department of Communications Engineering, University of the Basque Country (UPV/EHU), 48013 Bilbao, Spain
| | - Daniel Alonso
- Emergency Medical System (Emergentziak-Osakidetza), Basque Health Service, 20014 Donostia, Spain
| | - María del Mar Baqueriza
- Emergency Medical System (Emergentziak-Osakidetza), Basque Health Service, 20014 Donostia, Spain
| | - Karlos Ibarguren
- Emergency Medical System (Emergentziak-Osakidetza), Basque Health Service, 20014 Donostia, Spain
| |
Collapse
|
6
|
Figuera C, Irusta U, Morgado E, Aramendi E, Ayala U, Wik L, Kramer-Johansen J, Eftestøl T, Alonso-Atienza F. Machine Learning Techniques for the Detection of Shockable Rhythms in Automated External Defibrillators. PLoS One 2016; 11:e0159654. [PMID: 27441719 PMCID: PMC4956226 DOI: 10.1371/journal.pone.0159654] [Citation(s) in RCA: 40] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/03/2016] [Accepted: 07/05/2016] [Indexed: 01/08/2023] Open
Abstract
Early recognition of ventricular fibrillation (VF) and electrical therapy are key for the survival of out-of-hospital cardiac arrest (OHCA) patients treated with automated external defibrillators (AED). AED algorithms for VF-detection are customarily assessed using Holter recordings from public electrocardiogram (ECG) databases, which may be different from the ECG seen during OHCA events. This study evaluates VF-detection using data from both OHCA patients and public Holter recordings. ECG-segments of 4-s and 8-s duration were analyzed. For each segment 30 features were computed and fed to state of the art machine learning (ML) algorithms. ML-algorithms with built-in feature selection capabilities were used to determine the optimal feature subsets for both databases. Patient-wise bootstrap techniques were used to evaluate algorithm performance in terms of sensitivity (Se), specificity (Sp) and balanced error rate (BER). Performance was significantly better for public data with a mean Se of 96.6%, Sp of 98.8% and BER 2.2% compared to a mean Se of 94.7%, Sp of 96.5% and BER 4.4% for OHCA data. OHCA data required two times more features than the data from public databases for an accurate detection (6 vs 3). No significant differences in performance were found for different segment lengths, the BER differences were below 0.5-points in all cases. Our results show that VF-detection is more challenging for OHCA data than for data from public databases, and that accurate VF-detection is possible with segments as short as 4-s.
Collapse
Affiliation(s)
- Carlos Figuera
- Department of Telecommunication Engineering, Universidad Rey Juan Carlos, Madrid, Spain
- * E-mail:
| | - Unai Irusta
- Department of Communication Engineering, University of the Basque Country UPV/EHU, Bilbao, Spain
| | - Eduardo Morgado
- Department of Telecommunication Engineering, Universidad Rey Juan Carlos, Madrid, Spain
| | - Elisabete Aramendi
- Department of Communication Engineering, University of the Basque Country UPV/EHU, Bilbao, Spain
| | - Unai Ayala
- Electronics and Computing Department, University of Mondragon, Mondragon, Spain
| | - Lars Wik
- Norwegian National Advisory Unit on Prehospital Emergency Medicine (NAKOS), Oslo University Hospital and University of Oslo, Oslo, Norway
| | - Jo Kramer-Johansen
- Norwegian National Advisory Unit on Prehospital Emergency Medicine (NAKOS), Oslo University Hospital and University of Oslo, Oslo, Norway
| | - Trygve Eftestøl
- Department of Electrical Engineering and Computer Science, University of Stavanger, Stavanger, Norway
| | - Felipe Alonso-Atienza
- Department of Telecommunication Engineering, Universidad Rey Juan Carlos, Madrid, Spain
| |
Collapse
|
7
|
Requena-Carrión J, Alonso-Atienza F, Everss E, Sánchez-Muñoz JJ, Ortiz M, García-Alberola A, Rojo-Álvarez JL. Analysis of the robustness of spectral indices during ventricular fibrillation. Biomed Signal Process Control 2013. [DOI: 10.1016/j.bspc.2013.06.013] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
|
8
|
Requena-Carrión J, Beltrán-Molina FA, Marques AG. Relating the spectrum of cardiac signals to the spatiotemporal dynamics of cardiac sources. Biomed Signal Process Control 2013. [DOI: 10.1016/j.bspc.2013.09.002] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.1] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/26/2022]
|
9
|
Firoozabadi R, Nakagawa M, Helfenbein ED, Babaeizadeh S. Predicting defibrillation success in sudden cardiac arrest patients. J Electrocardiol 2013; 46:473-9. [DOI: 10.1016/j.jelectrocard.2013.06.007] [Citation(s) in RCA: 32] [Impact Index Per Article: 2.9] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/24/2013] [Indexed: 11/16/2022]
|
10
|
Wu X, Bisera J, Tang W. Signal integral for optimizing the timing of defibrillation. Resuscitation 2013; 84:1704-7. [PMID: 23969193 DOI: 10.1016/j.resuscitation.2013.08.005] [Citation(s) in RCA: 17] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/21/2013] [Revised: 08/01/2013] [Accepted: 08/12/2013] [Indexed: 12/01/2022]
Abstract
OBJECTIVE The possibility of successful defibrillation decreases with an increased duration of ventricular fibrillation (VF). Futile electrical shocks are inversely correlated with myocardial contractile function and long-term survival. Previous studies have demonstrated that various ECG waveform analyses predict the success of defibrillation. This study investigated whether the absolute amplitude of pre-shock VF waveform is likely to predict the success of defibrillation. METHODS ECG recordings of 350 out-of-hospital cardiac arrest (OOHCA) patients were obtained from the automated external defibrillator (AED) and analyzed by the method of signal integral. Successful defibrillation was defined as organized rhythm with heart rate ≥40beat/min commencing within one min of post-shock period and persisting for a minimum of 30s. RESULTS Signal integral was significantly greater in successful defibrillation than unsuccessful defibrillation (81.76±32.3mV vs. 34.9±15.33mV, p<0.001). The intersection of the sensitivity and specificity curve provided a threshold value of 51mV. The corresponding values of sensitivity, specificity, positive predictive and negative predictive values for successful defibrillation were 90%, 86%, 80% and 93%, respectively. The receiver operator curve further revealed that signal integral predicted the likelihood of successful defibrillation (area under the curve=0.949). CONCLUSIONS Signal integral predicted successful electrical shocks on patients with ventricular fibrillation and have potential to optimize the timing of defibrillation and reduce the number of electrical shocks.
Collapse
Affiliation(s)
- Xiaobo Wu
- Weil Institute of Critical Care Medicine, Rancho Mirage, CA, United States.
| | | | | |
Collapse
|
11
|
Neumar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, Kudenchuk PJ, Ornato JP, McNally B, Silvers SM, Passman RS, White RD, Hess EP, Tang W, Davis D, Sinz E, Morrison LJ. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010; 122:S729-67. [PMID: 20956224 DOI: 10.1161/circulationaha.110.970988] [Citation(s) in RCA: 888] [Impact Index Per Article: 63.4] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 12/14/2022]
Abstract
The goal of therapy for bradycardia or tachycardia is to rapidly identify and treat patients who are hemodynamically unstable or symptomatic due to the arrhythmia. Drugs or, when appropriate, pacing may be used to control unstable or symptomatic bradycardia. Cardioversion or drugs or both may be used to control unstable or symptomatic tachycardia. ACLS providers should closely monitor stable patients pending expert consultation and should be prepared to aggressively treat those with evidence of decompensation.
Collapse
|
12
|
|
13
|
Jekova I. Shock advisory tool: Detection of life-threatening cardiac arrhythmias and shock success prediction by means of a common parameter set. Biomed Signal Process Control 2007. [DOI: 10.1016/j.bspc.2007.01.002] [Citation(s) in RCA: 33] [Impact Index Per Article: 1.9] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/28/2022]
|
14
|
Abstract
PURPOSE OF REVIEW Ventricular fibrillation occurs during many cases of cardiac arrest and is treated with rescue shocks. Coarse ventricular fibrillation occurs earlier after the onset of cardiac arrest and is more likely to be converted to an organized rhythm with pulses by rescue shocks. Less organized or fine ventricular fibrillation occurs later, has less power concentrated within narrow frequency bands and lower amplitude, and is less likely to be converted to an organized rhythm by rescue shocks. Quantitative analysis of the ventricular fibrillation waveform may distinguish coarse ventricular fibrillation from fine ventricular fibrillation, allowing more appropriate delivery of rescue shocks. RECENT FINDINGS A variety of studies in animals and humans indicate that there is underlying structure within the ventricular fibrillation waveform. Highly organized or coarse ventricular fibrillation is characterized by large power contributions from a few component frequencies and higher amplitude. Amplitude, decomposition into power spectra, or probability-based, nonlinear measures all can quantify the organization of human ventricular fibrillation waveforms. Clinical data have accumulated that these quantitative measures, or combinations of these measures, can predict the likelihood of rescue shock success, restoration of circulation, and survival to hospital discharge. SUMMARY Many quantitative ventricular fibrillation measures could be implemented in current generations of monitors/defibrillators to assist the timing of rescue shocks during clinical care. Emerging data suggest that a period of chest compressions or reperfusion can increase the likelihood of successful defibrillation. Therefore, waveform-based prediction of defibrillation success could reduce the delivery of failed rescue shocks.
Collapse
Affiliation(s)
- Clifton W Callaway
- University of Pittsburgh, School of Medicine, Pittsburgh, Pennsylvania 15213, USA.
| | | |
Collapse
|