Atrial fibrillation occurrence risk prediction system based on heartbeat rhythm signals and application thereof

Abstract

The invention belongs to the field of physiological signal and electrocardiosignal analysis, and particularly relates to an atrial fibrillation occurrence risk prediction system based on heartbeat rhythm signals and application of the atrial fibrillation occurrence risk prediction system. In the prediction system, a heartbeat rhythm signal preprocessing module obtains heartbeat interval sequence data by calculating the time interval of two adjacent heartbeats; other complex feature extraction methods are not needed; in the sample labeling and model training module, during sample labeling, a positive sample is from heartbeat interval data during a sinus heart rate period when a paroxysmal atrial fibrillation patient does not suffer from atrial fibrillation, and a negative sample is from heartbeat interval data of a patient who does not suffer from atrial fibrillation in the future; and an atrial fibrillation risk prediction model is trained by using positive and negative samples and tests heartbeat interval data to be detected, a curve between the proportion of the number of future atrial fibrillation samples and probability thresholds is obtained according to different output probability thresholds, the area under the curve is calculated as an atrial fibrillation risk factor, and the factor can reflect the overall condition of the heartbeat interval data of a patient. Therefore, the occurrence risk of the atrial fibrillation can be accurately predicted.

Description

technical field [0001] The invention belongs to the field of analysis of physiological signals and electrocardiographic signals, and more specifically relates to a risk prediction system for occurrence of atrial fibrillation based on heartbeat rhythm signals and its application. Background technique [0002] Atrial fibrillation is the most common tachyarrhythmia disease, characterized by rapid and irregular beating of the atrium. About 33.5 million people worldwide suffer from atrial fibrillation. The incidence of atrial fibrillation is higher in the elderly population. Compared with women, the incidence of atrial fibrillation is higher in men high. Atrial fibrillation is one of the main factors leading to stroke, heart failure, and reduced quality of life and physical fitness. Undiagnosed and untreated atrial fibrillation may be a risk factor for dizziness, dyspnea, or stroke. Patients with atrial fibrillation may also induce sudden cardiac arrest. vascular disease. In ad...

AI Technical Summary

Problems solved by technology

[0004] At present, most of the intelligent algorithms related to atrial fibrillation are algorithms for identifying the onset of atrial fibrillation. These methods can detect the existence of atrial fibrillation only when the patient is experiencing atrial fibrillation; they cannot detect the presence of atrial fibrillation in the past However, patients with paroxysmal atrial fibrillation who did not have atrial fibrillation at the time of detection, or patients whose atrial fibrillation was still in its infancy and did not fully develop atrial fibrillation

However, there is evidence that in patients at risk of developing atrial fibrillation, even in normal ECG signals, due to changes in the structure of the atria, there will be some subtle changes, and by identifying such subtle changes It is possible to predict the risk of atrial fibrillation in advance before the onset of atrial fibrillation

[0005] Existing methods can only detect atrial fibrillation by analyzing the electrocardiogram recorded in the atrial fibrillation signal, or predict the risk of atrial fibrillation through the electrocardiogram waveform data, which has certain application limitations

Images