Method and device for synthesis of ECG V-lead signals from limb-lead measurement using R-peak aligned GAN

摘要

The present invention performs R peak alignment in the limb induction measurement signal of the electrocardiogram, and converts the limb induction measurement signal, which is a time domain signal, into a two-dimensional spatial signal through ordered time sequence embedding, and the previously learned generative hostile neural network (GAN) ), and outputs the chest induction signal of the electrocardiogram from the generative hostile neural network (GAN), without attaching multiple electrodes to the chest, using the generative hostile neural network (GAN) from the limb induction signal of the electrocardiogram. It relates to a method and apparatus for deriving a chest guidance signal. The present invention, a limb induction measuring device for measuring an electrocardiogram signal by limb induction; Equipped with a pre-learned artificial neural network, but when the electrocardiogram signal of limb guidance received from the limb guidance measuring device is input to the artificial neural network by pre-processing the signal for application of the artificial neural network, the artificial neural network outputs an electrocardiogram signal of chest induction. A chest induction electrocardiogram extraction device including an operation processing unit made; Including, characterized by a chest induction electrocardiogram reconstruction system. The artificial neural network is a GAN, and the artificial neural network input signal input to the artificial neural network is an electrocardiogram signal for limb guidance that is aligned with the R wave and converted into a bi-dimensional space.