Why the study?
Elevated troponin indicates cardiac injury, and rapid, reliable prediction of troponin elevation from ECGs could serve as a valuable time-saving diagnostic tool during emergency decision-making.
Does a residual convolutional neural network model accurately predict high-sensitivity troponin elevation from triage 12-lead ECGs in patients presenting to the emergency room with chest pain or dyspnoea?
Does a residual convolutional neural network model accurately predict high-sensitivity troponin elevation from triage 12-lead ECGs in patients presenting to the emergency room with chest pain or dyspnoea?
A deep learning ResNet model trained on 12-lead triage ECGs can predict high-sensitivity troponin elevation (incorporating sex- and assay-specific cut-offs) with a high negative predictive value of 86.6%, offering potential utility for rapid first-line triage of chest pain or dyspnoea in the emergency room.
Answers are grounded in this paper and cite their sources.
ECG-based deep learning may predict troponin elevation (NPV 86.6%); leaves open utility for triage or outcomes in prospective studies.
BACKGROUND: Elevated troponin levels are a sensitive biomarker for cardiac injury. The quick and reliable prediction of troponin elevation for patients with chest pain from readily available ECGs may pose a valuable time-saving diagnostic tool during decision-making concerning this patient population. METHODS AND RESULTS: The data used included 15 856 ECGs from patients presenting to the emergency rooms with chest pain or dyspnoea at two centres in Sweden from 2015 to June 2023. All patients had high-sensitivity troponin test results within 6 hours after 12-lead ECG. Both troponin I (TnI) and TnT were used, with biomarker-specific cut-offs and sex-specific cut-offs for TnI. On this dataset, a residual convolutional neural network (ResNet) was trained 10 times, each on a unique split of the data. The final model achieved an average area under the curve for the receiver operating characteristic curve of 0.7717 (95% CI±0.0052), calibration curve analysis revealed a mean slope of 1.243 (95% CI±0.075) and intercept of -0.073 (95% CI±0.034), indicating a good correlation between prediction and ground truth. Post-classification, tuned for F1 score, accuracy was 71.43% (95% CI±1.28), with an F1 score of 0.5642 (95% CI±0.0052) and a negative predictive value of 0.8660 (95% CI±0.0048), respectively. The ResNet displayed comparable or surpassing metrics to prior presented models. CONCLUSION: The model exhibited clinically meaningful performance, notably its high negative predictive accuracy. Therefore, clinical use of comparable neural networks in first-line, quick-response triage of patients with chest pain or dyspnoea appears as a valuable option in future medical practice.
No takes yet. Share an insight, caveat, or question.
Hilgendorf et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: