TY - JOUR
T1 - Convolution Neural Network Algorithm for Shockable Arrhythmia Classification Within a Digitally Connected Automated External Defibrillator
AU - Shen, Christine P.
AU - Freed, Benjamin C.
AU - Walter, David P.
AU - Perry, James C.
AU - Barakat, Amr F.
AU - Elashery, Ahmad Ramy A.
AU - Shah, Kevin S.
AU - Kutty, Shelby
AU - McGillion, Michael
AU - Ng, Fu Siong
AU - Khedraki, Rola
AU - Nayak, Keshav R.
AU - Rogers, John D.
AU - Bhavnani, Sanjeev P.
N1 - Funding Information:
Avive Solutions provided funding for this study. F.S.N. received the Programme Grant to Imperial College of London from the British Heart Foundation and grant funding from the National Institute for Health Research. J.C.P. received a grant from the National Institutes of Health R01.
Funding Information:
B.C.F. and D.P.W. were issued patent number US 11089989 B2 (45) with Avive Solutions. J.C.P. receives consultant fees from Protaryx and Alta Thera. M.M. is on the executive committee for the Society for Perioperative Care and receives patient monitoring equipment from Philips. F.S.N. receives funding from Programme Grant to Imperial College of London and National Institute for Heath and Care Research grant funding. S.P.B. received consulting fees from Bristol Meyers Squibb, Pfizer, and Infinion; participated on the advisory board for Proteus Digital; had a leadership role in the American College of Cardiology, American Society of Echocardiography, and Biocom (all positions unpaid and voluntary).
Funding Information:
The AHA, with funding from the National Heart, Lung, and Blood Institute (NHLBI), developed a database for evaluation of ventricular arrhythmia detectors during the
Publisher Copyright:
© 2023 The Authors.
PY - 2023/4/18
Y1 - 2023/4/18
N2 - BACKGROUND: Diagnosis of shockable rhythms leading to defibrillation remains integral to improving out-of-hospital cardiac arrest outcomes. New machine learning techniques have emerged to diagnose arrhythmias on ECGs. In out-of-hospital cardiac arrest, an algorithm within an automated external defibrillator is the major determinant to deliver defibrillation. This study developed and validated the performance of a convolution neural network (CNN) to diagnose shockable arrhythmias within a novel, miniaturized automated external defibrillator. METHODS AND RESULTS: There were 26 464 single-lead ECGs that comprised the study data set. ECGs of 7-s duration were retrospectively adjudicated by 3 physician readers (N=18 total readers). After exclusions (N=1582), ECGs were divided into training (N=23 156), validation (N=721), and test data sets (N=1005). CNN performance to diagnose shockable and non-shockable rhythms was reported with area under the receiver operating characteristic curve analysis, F1, and sensitivity and specificity calculations. The duration for the CNN to output was reported with the algorithm running within the automated external defibrillator. Internal and external validation analyses included CNN performance among arrhythmias, often mistaken for shockable rhythms, and performance among ECGs modified with noise to mimic artifacts. The CNN algorithm achieved an area under the receiver operating characteristic curve of 0.995 (95% CI, 0.990–1.0), sensitivity of 98%, and specificity of 100% to diagnose shockable rhythms. The F1 scores were 0.990 and 0.995 for shockable and nonshockable rhythms, respectively. After input of a 7-s ECG, the CNN generated an output in 383±29 ms (total time of 7.383 s). The CNN outperformed adjudica-tors in classifying atrial arrhythmias as nonshockable (specificity of 99.3%– 98.1%) and was robust against noise artifacts (area under the receiver operating characteristic curve range, 0.871– 0.999). CONCLUSIONS: We demonstrate high diagnostic performance of a CNN algorithm for shockable and nonshockable rhythm ar-rhythmia classifications within a digitally connected automated external defibrillator. REGISTRATION: URL: https://clinicaltrials.gov/ct2/show/NCT03662802; Unique identifier: NCT03662802.
AB - BACKGROUND: Diagnosis of shockable rhythms leading to defibrillation remains integral to improving out-of-hospital cardiac arrest outcomes. New machine learning techniques have emerged to diagnose arrhythmias on ECGs. In out-of-hospital cardiac arrest, an algorithm within an automated external defibrillator is the major determinant to deliver defibrillation. This study developed and validated the performance of a convolution neural network (CNN) to diagnose shockable arrhythmias within a novel, miniaturized automated external defibrillator. METHODS AND RESULTS: There were 26 464 single-lead ECGs that comprised the study data set. ECGs of 7-s duration were retrospectively adjudicated by 3 physician readers (N=18 total readers). After exclusions (N=1582), ECGs were divided into training (N=23 156), validation (N=721), and test data sets (N=1005). CNN performance to diagnose shockable and non-shockable rhythms was reported with area under the receiver operating characteristic curve analysis, F1, and sensitivity and specificity calculations. The duration for the CNN to output was reported with the algorithm running within the automated external defibrillator. Internal and external validation analyses included CNN performance among arrhythmias, often mistaken for shockable rhythms, and performance among ECGs modified with noise to mimic artifacts. The CNN algorithm achieved an area under the receiver operating characteristic curve of 0.995 (95% CI, 0.990–1.0), sensitivity of 98%, and specificity of 100% to diagnose shockable rhythms. The F1 scores were 0.990 and 0.995 for shockable and nonshockable rhythms, respectively. After input of a 7-s ECG, the CNN generated an output in 383±29 ms (total time of 7.383 s). The CNN outperformed adjudica-tors in classifying atrial arrhythmias as nonshockable (specificity of 99.3%– 98.1%) and was robust against noise artifacts (area under the receiver operating characteristic curve range, 0.871– 0.999). CONCLUSIONS: We demonstrate high diagnostic performance of a CNN algorithm for shockable and nonshockable rhythm ar-rhythmia classifications within a digitally connected automated external defibrillator. REGISTRATION: URL: https://clinicaltrials.gov/ct2/show/NCT03662802; Unique identifier: NCT03662802.
KW - ECG
KW - automated external defibrillator
KW - convolution neural network
KW - machine learning
KW - ventricular arrhythmias
UR - http://www.scopus.com/inward/record.url?scp=85152635908&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85152635908&partnerID=8YFLogxK
U2 - 10.1161/JAHA.122.026974
DO - 10.1161/JAHA.122.026974
M3 - Article
C2 - 36942628
AN - SCOPUS:85152635908
SN - 2047-9980
VL - 12
JO - Journal of the American Heart Association
JF - Journal of the American Heart Association
IS - 8
M1 - e026974
ER -