Model establishment method and system for extracting time-specific data of ECG

Abstract

The invention relates to a model establishment method and system for extracting time-specific data of ECG. The model establishment method includes the steps that S1, DCG information of sudden cardiacdeath survivors and DCG information of normal subjects are obtained; S2, obtained ECG signals are de-noised by using wavelet de-noising to lower power frequency interference with maximum frequency andminimum period and baseline drift with maximum period and minimum frequency; S3, the peak of ECG data after de-noising is determined and characteristic parameters are extracted; S4, logistic regression models are established for time points according to peak determination and characteristic parameter extraction results, the characteristic parameters are selected, and the characteristic parametersof the time points are selected by five times crossover operation; and S5, according to the selected characteristic parameters, the models of the points are synthesized through a Bayes theory, and afinal model is established. According to the model establishment method and system for extracting the time-specific data of the ECG, time-specific data extraction of the ECG signals of sudden cardiacdeath at 24 hours can be realized.

Description

technical field [0001] The invention relates to the field of medical signal processing, in particular to a model building method and system for extracting electrocardiogram time-specific data. Background technique [0002] Electrocardiogram refers to the graph that records the electrical signals generated by the periodic activity of the heart and draws them in a prescribed format. The electrocardiogram reflects the electrical activity process of the heart's excitement. It has important reference value for the basic functions of the heart and its pathological research. With its advantages of economical, reliable, fast and non-invasive measurement methods, electrocardiogram has been widely used in clinical practice. The electrophysiological activity of the heart can be directly reflected on the electrocardiogram. The waveform, cycle and other information contained in the electrocardiogram have important reference value for doctors . [0003] Sudden cardiac death refers to nat...

AI Technical Summary

Problems solved by technology

[0007] Most of the ECG early warning signals in the above method come from a single ECG measurement, ignoring the time-specific dynamic change information of the 24-hour ECG in patients with sudden cardiac death, and, in terms of program identification, although the existing technology is relatively mature for the detection of R waves , but there is a lack of extraction and use of P, Q, S, T peak positioning and peak area and other related data. The specificity and sensitivity of these ECG early warning signals used in the past are low.

Images