Pulse Anomaly Detection Algorithm And Example Verification Based On Random Matrix

With the development of society,people’s various unhealthy living and eating habits may lead to the occurrence of cardiovascular disease.The early stage of cardiovascular disease is hidden and difficult to be detected by people,but the onset is often very rapid.Electrocardiography has a long history and is widely used in diagnosing cardiovascular diseases,but the use of electrocardiographs to collect electrocardiograms is complicated and requires high professional requirements.Traditional Chinese medicine believes that pulse contains a lot of information,which can be used to diagnose cardiovascular diseases,but it has not been widely promoted due to the characteristics of pulse diagnosis relying on doctors’ experience and insufficient medical resources.In view of this situation,combined with the development trend of close-fitting,compact and real-time medical monitoring,this paper proposes a pulse short-term window abnormality judgment algorithm based on random matrix,and builds a real-time flow pulse abnormality calculation platform to judge whether the pulse is real-time.Abnormal,the main research contents are as follows:1.For the real-time pulse,a short-time window pulse abnormality calculation algorithm based on random matrix and a quartile outlier judgment algorithm suitable for real-time flow are proposed.The random matrix algorithm calculates the linear characteristic statistics,and the quartile algorithm judges whether the Exception,such algorithms are designed with real-time and specificity in mind2.Use the ECG signal and pulse signal measured at the same time in the network public database,use the algorithm introduced in the literature to identify the abnormality of the ECG signal,use the random matrix algorithm proposed in this paper to identify the abnormality of the pulse signal,and the matching degree of the two abnormal results was 75.14%.Matching degree,and experimentally explore the influence of external disturbances on the test results during the test pulse process.3.Combined with the characteristics of wearable and real-time monitoring of medical equipment,a real-time flow pulse abnormality calculation system is built,and the user’s pulse is tested with a finger clip.The random matrix algorithm and the quartile algorithm provide real-time pulse abnormality judgment results in a short-term window.The performance of the real-time stream computing system is tested.The overall delay of the system is 4.942s,which has good real-time performanceThe research content of this paper can be used in real-time monitoring and abnormal judgment of pulse signal,and has a relatively broad application prospect.