Feature fusion transfer learning arrhythmia classification system based on 2D heart beat

Abstract

The invention discloses a feature fusion transfer learning arrhythmia classification system based on 2D heart beat, and the system comprises: an obtaining module which is configured to obtain a targetelectrocardiosignal; the reconstruction module that is configured to reconstruct the target electrocardiosignal from one dimension to a two-dimensional electrocardiosignal; the feature extraction module that is configured to respectively extract a first feature and a second feature from the two-dimensional electrocardiosignal; the feature fusion module that is configured to perform fusion processing on the first feature and the second feature; and the classification module that is configured to input the fused features into the trained classifier and output an arrhythmia classification resultcorresponding to the current target electrocardiosignal.

Description

technical field [0001] The present application relates to the technical field of heart rate duration classification system, in particular to a 2D heartbeat-based feature fusion transfer learning arrhythmia classification system. Background technique [0002] The statements in this section merely mention the background art related to this application, and do not necessarily constitute the prior art. [0003] There are two main types of traditional arrhythmia classification methods, one is feature extraction combined with machine learning methods, and the other is deep learning methods without feature extraction. [0004] Machine learning method based on feature extraction: Traditional machine learning algorithms need to manually extract features including time domain, frequency domain, etc. based on prior experience, and then train an appropriate machine learning classifier. Therefore, the advantages and disadvantages of the extracted features affect the final classification...

AI Technical Summary

Problems solved by technology

Sayantan G et al. use deep belief network and active learning to classify ECG beats, but this method only uses one-dimensional Gauss-Bernoulli deep belief network to learn the feature representation of ECG

Shi et al. proposed a novel multi-input deep learning network and used it for atrial fibrillation detection, however, this method based on active and transfer learning is still based on one-dimensional ECG data

In fact, since the collected ECG data is in one-dimensional form, the above-mentioned 1-D CNN method has the advantages of fast running speed and simple structure in the classification of ECG signal arrhythmia. The information contained is limited and cannot fully reflect all the information of cardiac physiological changes contained in the data, so the classification accuracy of this 1-D CNN method is difficult to further improve

Mashrur et al. used wavelet transform to convert one-dimensional ECG signals into two-dimensional time-frequency images, and then used AlexNet convolutional neural network to learn to recognize atrial fibrillation. However, the wavelet transform still obtained narrow-band signals, and single frequency information could not be obtained accurately.

Images