Method for detecting electrocardio waveform characteristics based on wavelet transform and logistic regression algorithm

Abstract

The invention discloses a method for detecting electrocardio waveform characteristics based on a wavelet transform and logistic regression algorithm. Except for the waveform range, a wider characteristic is applied to judge if one waveform is a real QSR; these characteristics are energy value within a specific frequency scale after wavelet transform, interval with previous suspected QRS wave, width of suspected QPS wave, voltage change rate of suspected QRS wave; every suspected QRS wave is compared with an expert marked QRS wave to judge if they are real QRS waveform; a characteristic matrixacquired in the previous step is input to a logic regression model, the gradient descent is the optimized method; a learning rate is set as 0.02 and iterated until convergence; the training adopts themethod of 1-remained intersection verification. The algorithm can acquire authority of every characteristic parameter from the massive marked patient data, thereby avoiding the demand for a manual definition rule; the algorithm can identify the QRS wave group while acquire reliable signal quality information; thus the method is helpful to reduce the wrong alarm.

Description

technical field [0001] The invention relates to the field of medical technology, in particular to a method for detecting electrocardiographic waveform features, in particular to a method for detecting electrocardiographic waveform features based on wavelet transform and logistic regression algorithm. Background technique [0002] The myocardial cell membrane is a semipermeable membrane. In the resting state, a certain number of positively charged cations are arranged outside the membrane, and the same number of negatively charged anions are arranged inside the membrane. The potential outside the membrane is higher than that inside the membrane, which is called a polarized state. In the resting state, since the cardiomyocytes in all parts of the heart are in a polarized state, there is no potential difference, and the potential curve traced by the current recorder is straight, which is the equipotential line of the body surface ECG. When cardiomyocytes are stimulated by a cer...

AI Technical Summary

Problems solved by technology

[0008] 1. The discrimination of QRS wave mainly depends on the waveform amplitude. Therefore, when the T wave amplitude is large, it is easy to be mistakenly detected as QRS wave, and it is sensitive to noise interference. When the signal quality is poor, the detection accuracy of QRS wave will drop significantly;

[0009] 2. Selecting the threshold through artificially defined rules leads to insufficient robustness, especially when the error rate is easy to increase when analyzing patients with rare arrhythmias

Images