Diagnostic accuracy of prehospital electrocardiograms interpreted remotely by emergency physicians in myocardial infarction patients

Tanguay, Alain; J, Lebon; Brassard, Eric; Hébert, Denise; Bégin, François

doi:10.1016/j.ajem.2018.09.012

Cited by 7 publications

(8 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our model’s misclassifications as STEMI follows known clinical and machine learning patterns, with perimyocarditis as an important impostor 26 . The conditions over-represented in those misclassified as NSTEMI were logical to some extent, such as valvular or congenital heart disease and pulmonary edema; the gastric ulcer and dental traits more surprising.…”

Section: Discussionmentioning

confidence: 89%

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Gustafsson

Gedon

Lampa

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Myocardial infarction diagnosis is a common challenge in the emergency department. In managed settings, deep learning-based models and especially convolutional deep models have shown promise in electrocardiogram (ECG) classification, but there is a lack of high-performing models for the diagnosis of myocardial infarction in real-world scenarios. We aimed to train and validate a deep learning model using ECGs to predict myocardial infarction in real-world emergency department patients. We studied emergency department patients in the Stockholm region between 2007 and 2016 that had an ECG obtained because of their presenting complaint. We developed a deep neural network based on convolutional layers similar to a residual network. Inputs to the model were ECG tracing, age, and sex; and outputs were the probabilities of three mutually exclusive classes: non-ST-elevation myocardial infarction (NSTEMI), ST-elevation myocardial infarction (STEMI), and control status, as registered in the SWEDEHEART and other registries. We used an ensemble of five models. Among 492,226 ECGs in 214,250 patients, 5,416 were recorded with an NSTEMI, 1,818 a STEMI, and 485,207 without a myocardial infarction. In a random test set, our model could discriminate STEMIs/NSTEMIs from controls with a C-statistic of 0.991/0.832 and had a Brier score of 0.001/0.008. The model obtained a similar performance in a temporally separated test set of the study sample, and achieved a C-statistic of 0.985 and a Brier score of 0.002 in discriminating STEMIs from controls in an external test set. We developed and validated a deep learning model with excellent performance in discriminating between control, STEMI, and NSTEMI on the presenting ECG of a real-world sample of the important population of all-comers to the emergency department. Hence, deep learning models for ECG decision support could be valuable in the emergency department.

show abstract

Section: Discussionmentioning

confidence: 89%

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Gustafsson

Gedon

Lampa

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…11,13,14,33 Our study is the largest to examine the appropriateness of CCL activation utilizing PH-ECG transmission to interventional cardiologists. In previous studies, the PH-ECG was transmitted to emergency physicians, [17][18][19][20] and in two studies, cardiologists were part of the ECG reading team 34 or consulted in unclear cases. 13 Initial studies on PH-ECG transmission conducted more than a decade ago primarily examined the effect of PH-ECG transmission on reperfusion times, and compared transmission to cardiologists with historical controls, self-presenting patients, or failed transmissions.…”

Section: Resultsmentioning

confidence: 99%

“…3,4 The small of number studies comparing the accuracy of PH-ECG computer interpretation (with or without paramedic over-read) with transmission for clinician interpretation have reported conflicting results. 13,17,18 In addition, these studies, 13,17,18 and others 19,20 reporting on the accuracy of PH-ECG transmission primarily involved transmission to emergency physicians for over-read. As examination of the appropriateness of CCL activation using PH-ECG transmission with interventional cardiologist over-read has not been assessed, we evaluated this strategy.…”

Section: Introductionmentioning

confidence: 99%

Utility of prehospital electrocardiogram interpretation in ST‐segment elevation myocardial infarction utilizing computer interpretation and transmission for interventional cardiologist consultation

Faour

Cherrett

Gibbs

et al. 2022

Cathet Cardio Intervent

View full text Add to dashboard Cite

Objectives: We examined the appropriateness of prehospital cardiac catheter laboratory activation (CCL-A) in ST-segment elevation myocardial infarction (STEMI) utilizing the University of Glasgow algorithm (UGA) and remote interventional cardiologist consultation. Background:The incremental benefit of prehospital electrocardiogram (PH-ECG) transmission on the diagnostic accuracy and appropriateness of CCL-A has been examined in a small number of studies with conflicting results. Methods:We identified consecutive PH-ECG transmissions between June 2, 2010 and October 6, 2016. Blinded adjudication of ECGs, appropriateness of CCL-A, and index diagnoses were performed using the fourth universal definition of MI. The primary outcome was the appropriate CCL-A rate. Secondary outcomes included rates of false-positive CCL-A, inappropriate CCL-A, and inappropriate CCL nonactivation.Results: Among 1088 PH-ECG transmissions, there were 565 (52%) CCL-As and 523 (48%) CCL nonactivations. The appropriate CCL-A rate was 97% (550 of 565 CCL-As), of which 4.9% (n = 27) were false-positive. The inappropriate CCL-A rate was 2.7% (15 of 565 CCL-As) and the inappropriate CCL nonactivation rate was 3.6% (19 of 523 CCL nonactivations). Reasons for appropriate CCL nonactivation (n = 504) included nondiagnostic ST-segment elevation (n = 128, 25%), bundle branch block (n = 132, 26%), repolarization abnormality (n = 61, 12%), artefact (n = 72, 14%), no ischemic symptoms (n = 32, 6.3%), severe comorbidities (n = 26, 5.2%), transient ST-segment elevation (n = 20, 4.0%), and others.

show abstract

“…Our model's misclassi cations as STEMI follows known clinical and machine learning patterns, with perimyocarditis as an important impostor. 30 The conditions over-represented in those misclassi ed as NSTEMI were logical to some extent, such as valvular or congenital heart disease and pulmonary edema; the gastric ulcer and dental traits more surprising.…”

Section: Discussionmentioning

confidence: 99%

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Gustafsson

Gedon

Lampa

et al. 2022

Preprint

View full text Add to dashboard Cite

Myocardial infarction diagnosis is a common challenge in the emergency department. In managed settings, deep learning-based models and especially convolutional deep models have shown promise in electrocardiogram (ECG) classification, but there is a lack of high-performing models for the diagnosis of myocardial infarction in real-world scenarios. We aimed to train and validate a deep learning model using ECGs to predict myocardial infarction in real-world emergency department patients. We studied emergency department patients in the Stockholm region between 2007 and 2016 that had an ECG obtained because of their presenting complaint. We developed a deep neural network based on convolutional layers similar to a residual network. Inputs to the model were ECG tracing, age, and sex; and outputs were the probabilities of three mutually exclusive classes: non-ST-elevation myocardial infarction (NSTEMI), ST-elevation myocardial infarction (STEMI), and control status, as registered in the SWEDEHEART and other registries. We used an ensemble of five models. Among 492,226 ECGs in 214,250 patients, 5,416 were recorded with an NSTEMI, 1,818 a STEMI, and 485,207 without a myocardial infarction. In a random test set, our model could discriminate STEMIs/NSTEMIs from controls with a C-statistic of 0.991/0.832 and had a Brier score of 0.001/0.008. The model obtained a similar performance in a temporally separated test set, and achieved a C-statistic of 0.985 and a Brier score of 0.002 in discriminating STEMIs from controls in an external test set. We developed and validated a deep learning model with excellent performance in discriminating between control, STEMI, and NSTEMI on the presenting ECG of a real-world sample of the important population of all-comers to the emergency department. Hence, deep learning models for ECG decision support could be valuable in the emergency department.

show abstract

Diagnostic accuracy of prehospital electrocardiograms interpreted remotely by emergency physicians in myocardial infarction patients

Cited by 7 publications

References 24 publications

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Utility of prehospital electrocardiogram interpretation in ST‐segment elevation myocardial infarction utilizing computer interpretation and transmission for interventional cardiologist consultation

Development and validation of deep learning ECG-based prediction of myocardial infarction in emergency department patients

Contact Info

Product

Resources

About