Electrocardiosignal atrial fibrillation detection device based on dense connection convolutional recurrent neural network

Abstract

The invention discloses an electrocardiosignal atrial fibrillation detection device based on a dense connection convolutional recurrent neural network. The electrocardiosignal atrial fibrillation detection device comprises a data acquisition module, a preprocessing module, an atrial fibrillation detection module and a training module; the atrial fibrillation detection module is used for building adense connection convolutional recurrent neural network atrial fibrillation detection model; the model comprises a convolution layer, and a dense connection neural network, a bidirectional recurrentneural network and an output discrimination classification layer. The dense connection convolutional neural network effectively solves the problems of gradient disappearance and network ductility, and makes full use of characteristics at the same time; the bidirectional recurrent neural network enables the network to be more suitable for an analysis scene of time sequence signals; and the detection device detects electrocardiosignals from a space domain and a time domain successively, and considerable atrial fibrillation detection accuracy is finally achieved through combination and cascadingof the two networks. According to the technical scheme, compared with traditional atrial fibrillation segmented detection processes, the operation process is simpler, and robustness and algorithm stability are higher.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence deep neural network and electrocardiographic signal recognition, and in particular relates to a device for detecting atrial fibrillation of electrocardiographic signals based on a densely connected convolutional cyclic neural network. Background technique [0002] Atrial fibrillation, referred to as atrial fibrillation, is a relatively common clinical arrhythmia disease with high potential harm. Although it is not a fatal disease, it is mainly due to the electrophysiological changes caused by it can cause discomfort symptoms such as fatigue, palpitations and chest pain in patients. Complications such as cerebral infarction and high blood pressure may affect life safety and health. It is precisely because of its early concealment and serious long-term disease effects that it is necessary to have an efficient and accurate detection algorithm for atrial fibrillation. The gold standa...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the deficiencies in the above-mentioned prior art, the present invention proposes an ECG signal atrial fibrillation detection device based on a densely connected convolutional neural network, which simplifies the complex feature extraction and preprocessing process of the traditional method, and realizes the use of end-to-end Atrial fibrillation detection algorithm based on end neural network, and proposed a method for detection of atrial fibrillation in ECG signals based on densely connected convolutional cyclic neural network. Connected Convolutional Neural Networks Applied to One-Dimensional Time-Series ECG Signal Analysis

Images