Beat-by-beat heart rate detection method based on multi-example learning and evolutionary optimization

Abstract

The invention discloses a beat-by-beat heart rate detection method based on multi-example learning and evolutionary optimization. The beat-by-beat heart rate detection method mainly solves the problems of high dependence on manual marking and low estimation accuracy of the heart rate of a BCG signal in the prior art. According to a scheme of the invention, the beat-by-beat heart rate detection method comprises the following steps: acquiring an original ballistocardiogram signal and a finger electric signal; extracting heartbeat signal characteristics of the ballistocardiogram signal, dividingthe heartbeat signal characteristics into positive packets and negative packets, dividing the positive packets and negative packets into a training set and a test set, learning the training sample setso as to obtain an initialized heartbeat template signal, and performing dimensionality reduction on the initialized heartbeat template signal so as to obtain a dimensionality-reduced heartbeat template signal; performing iterative optimization on the dimensionality-reduced heartbeat template signal so as to obtain an optimal heartbeat template signal; and performing classification detection on the test sample set by utilizing the optimal heartbeat template signal so as to obtain a final heart rate detection result. The beat-by-beat heart rate detection method provided by the invention improves the estimation accuracy of the heart rate of the BCG signal, has low requirement on heartbeat feature initialization, is low in manual marking cost, and can be used for heartbeat detection of imprecise marked ballistocardiogram signals.

Description

technical field [0001] The invention belongs to the technical field of medical signal processing, and further relates to a method for estimating heart rate, which can be used for heartbeat detection of imprecisely marked ballistocardiogram signals. Background technique [0002] Heart rate is one of the important vital signs to evaluate the health status of people, especially those with cardiovascular disease. Most heart disease patients need to take medicine for life, even if the clinical symptoms disappear, heart problems may recur at any time. Real-time heart rate monitoring is essential to prevent heart disease. The methods currently used to measure heart rate are: electrocardiography ECG, optical plethysmography PPG, phonocardiography PCG and ballistocardiography BCG. Each method determines heart rate by measuring different phenomena that occur in the body during the heartbeat, or heartbeat cycle. [0003] Compared with ECG, BCG records the ballistocardiogram signal i...

AI Technical Summary

Problems solved by technology

However, the disadvantage of this method is that: to obtain the same good classification accuracy for different data, a large number of feature extraction and feature selection are required, which obviously greatly increases the workload of researchers, and the difference in feature selection will affect the method. have a certain impact on the classification results of

Images