Research On Monitoring And Processing Technology Of ECG Signal Of Bedridden Patients

Nowadays,cardiovascular disease has seriously threatened people’s life and health.For bedridden patients,the problem is even more serious.Due to the problem of bedridden,when cardiovascular disease occurs,the patient cannot take measures such as calling for help in time;Vascular problems are a real concern for bedridden patients.The ECG signal of bedridden patients has two main characteristics due to the reasons such as the patient’s bedridden and the disease itself: first,the noise doped in the ECG signal of the bedridden patient is more than that of the general ECG signal;Second,the type and quantity of abnormal ECG signals in bedridden patients are also more complex than those of normal people.Based on this,this paper proposes an improved ECG signal denoising method and builds a fusion ECG signal classification and recognition model.,and introduced the MIT-BIH database as the basic data set for testing the effect of the improved algorithm.The MIT-BIH database is currently an internationally recognized arrhythmia database,of which 60% of the data comes from inpatients,which is suitable as the data set for this study.Finally,the monitoring of the electrocardiogram(ECG)of the bedridden patient is realized through the hardware design part.The processing of ECG signals of bedridden patients mainly includes three steps:ECG signal denoising,ECG signal feature extraction,and ECG signal classification.At present,people have carried out a lot of research on the denoising and classification of ECG signals.However,due to the complex noise contained in ECG signals,and the types and quantities of abnormal ECG signals are divided,the current ECG signal denoising and classification The method failed to achieve the desired effect.Based on this,this paper mainly focuses on the following three parts:1.Study an improved method for denoising of ECG signal: This paper proposes an improved denoising method for ECG signal for bedridden patients with weak ECG signal,low frequency range,more noise and incomplete denoising.-Wavelet thresholding method based on optimized variational modal decomposition(VMD).The MIT-BIH database is selected as the data set of ECG signals.On this basis,two other denoising methods are used for comparison.The results show that the method proposed in this paper has better denoising effect and is more conducive to subsequent ECG signals.Feature extraction and classification.2.Construct a fused ECG signal classification model: In this paper,the type and quantity of abnormal signals in the ECG signals of bedridden patients are large,and the current ECG signal classification methods have low accuracy and weak generalization ability.An ECG signal classification method based on a fusion attention mechanism-convolutional neural network(CBAM-CNN)and support vector machine(SVM)model.Using the MIT-BIH ECG database as the data set for the classification effect test,and compared with several other ECG signal classification methods,the evaluation parameters such as accuracy show that the hybrid model proposed in this paper is more suitable for ECG signal classification.3.Hardware design part,the hardware part of this paper is mainly the design of the ECG monitoring system for bedridden patients,the acquisition of ECG signals of bedridden patients is realized through hardware design,and the upper computer part displays the monitoring results and saves them.