AI-Enhanced Reconstruction of the 12-Lead Electrocardiogram via 3-Leads with Accurate Clinical Assessment

Federico Mason,Amitabh C. Pandey,Matteo Gadaleta,Eric J. Topol,Evan D. Muse,Giorgio Quer

DOI: https://doi.org/10.1101/2024.01.30.24302001

2024-01-30

Abstract:The 12-lead electrocardiogram (ECG) is an integral component to the diagnosis of a multitude of cardiovascular conditions. It is performed using a complex set of skin surface electrodes, limiting its use outside traditional clinical settings. We developed an artificial intelligence algorithm, trained over 600,000 clinically acquired ECGs, to explore whether fewer leads as input are sufficient to reconstruct a full 12-lead ECG. Two limb leads (I and II) and one precordial lead (V3) were required to generate a reconstructed synthetic 12-lead ECG highly correlated with the original ECG. An automatic algorithm for detection of acute myocardial infarction (MI) performed similarly for original and reconstructed ECGs (AUC=0.94). When interpreted by cardiologists, reconstructed ECGs achieved an accuracy of 81.4±5.0% in identifying ST elevation MI, comparable with the original 12-lead ECGs (accuracy 84.6±4.6%). These results will impact development efforts to innovate ECG acquisition methods with simplified tools in non-specialized settings.

Cardiovascular Medicine

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to use fewer leads (only two limb leads and one precordial lead) to reconstruct a complete 12 - lead electrocardiogram (ECG). Specifically, the researchers developed an artificial intelligence algorithm, aiming to explore whether a synthetic 12 - lead ECG highly correlated with the original 12 - lead ECG can be generated through the input of three leads (two limb leads I and II, and one precordial lead V3). The application of this technology can simplify the ECG collection method, enabling it to be used in non - professional environments, thereby improving the diagnostic efficiency and accessibility of cardiovascular diseases.

### Main problems:
1. **Reducing the number of leads**: The traditional 12 - lead ECG requires the use of a complex skin - surface electrode array, which limits its application outside the traditional clinical environment. Through the development of an AI algorithm, this study attempts to use fewer leads to reconstruct a complete 12 - lead ECG, in order to simplify the ECG collection process.
2. **Clinical evaluation**: The study also evaluated the clinical performance of the reconstructed 12 - lead ECG in the detection of acute myocardial infarction (AMI), especially the ability to identify ST - segment elevation myocardial infarction (STEMI).

### Research methods:
- **Data set**: The study used more than 600,000 clinically obtained ECG data for training.
- **Algorithm design**: An AI algorithm based on the Residual Convolutional Neural Network (RCNN) was developed, which can reconstruct a complete 12 - lead ECG from three leads (I, II, and V3).
- **Performance evaluation**: The performance of the reconstruction algorithm was evaluated by the Mean Squared Error (MSE) and the Coefficient of Determination (R²), and the clinical validity of the reconstructed ECG was verified using an automatic detection algorithm and the interpretation of cardiologists.

### Key findings:
- **Reconstruction performance**: When using leads I, II, and V3 as inputs, the reconstructed 12 - lead ECG is highly correlated with the original ECG, with an MSE of 0.0126 ± 0.0001 mV² and an R² of 72.62 ± 0.13%.
- **Clinical application**: The reconstructed ECG performs well in the automatic detection of AMI and the interpretation of STEMI by cardiologists, with accuracies of 81.4 ± 5.0% and 84.6 ± 4.6% respectively, which are comparable to those of the original 12 - lead ECG.

### Conclusion:
This study demonstrates the feasibility of reconstructing a complete 12 - lead ECG from a limited number of leads through an AI algorithm, which not only improves the application potential of ECG in non - professional environments but may also accelerate the diagnosis and treatment of acute myocardial infarction in emergency situations.