Parameters Optimization of Convolutional Neural Network to Synthesize 12-Lead ECG from a 3-Lead

This paper presents a metaheuristic optimization-based approach for reconstructing a 12-lead electrocardiogram (ECG) from a reduced set of leads using a convolutional neural network (CNN) model. The proposed method has two parts. The first part illustrates the optimization of the hyperparameters of the CNN model using the Particle Swarm Optimization (PSO) metaheuristic. The second part presents the optimization of the CNN model using a subset of the PTB diagnostic ECG database. The proposed deep learning model will be evaluated using the two-performance metrics, namely correlation coefficient (r), and root mean square error (RMSE). The results will be detailed and compared with traditional linear regression and LSTM deep learning model.